awesome-free-ai-coding

# awesome-free-ai-coding

> Legal free LLM APIs and coding agents for AI coding — free tiers, no-card trials and free models, probe-verified twice a week against live model catalogs and pricing pages. Generated 2026-09-27 from the registry; every offer above the Archived heading passed a probe in the last 60 days and has not failed 3 in a row.

Each offer links to a page with the free tier in the vendor's own words, the connection details (base URL, where to get a key, model ids, an Anthropic-format URL where the vendor documents one), the evidence the probe reads and the row's history. "No card" means the vendor asks for no payment method; "no key" means the endpoint answers without an account; "no account" means the vendor prints a key anyone may call it with. Offers the list carried and carries no more are listed last, under Archived, each with why it left.

## Coding agents & CLIs

- [opencode](https://mvalentsev.github.io/awesome-free-ai-coding/providers/opencode/): Open-source coding agent whose opencode Zen gateway prices a rotating set of models at zero — inside OpenCode only, no sign-in; any provider via BYOK; no card; what you send may be used to train or improve models; free models: `big-pickle`, `mimo-v2.5`, `ling-3.0-flash-fin`, `nemotron-3-ultra`, `nemotron-3.5-lightning`, `muse-spark-1.3-contributor`
- [Kilo Code](https://mvalentsev.github.io/awesome-free-ai-coding/providers/kilo-code/): Open-source VS Code / JetBrains / CLI agent whose $0 plan routes "Auto Free" to the models the Kilo Gateway marks free; the same gateway serves them to any OpenAI client without a key, with BYOK and local models alongside; no card; what you send may be used to train or improve models; no key; OpenAI-compatible at https://api.kilo.ai/api/gateway; Codex CLI profile at https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/configs/codex/kilo-code.config.toml; free models: `nemotron-3-ultra`, `nemotron-3-super`, `north-mini-code`, `step-3.7-flash`, `laguna-s-2.1`, `laguna-xs-2.1`, `nemotron-3.5-lightning`, `inkling-small`, `dots-3-note`, `ling-3.0-flash-fin`, `ling-3.0-flash-sante`, `nemotron-3-nano-omni`
- [Google Antigravity](https://mvalentsev.github.io/awesome-free-ai-coding/providers/antigravity/): Google's agent-first IDE and CLI, and where the Gemini CLI free tier went — Gemini CLI and the Code Assist IDE extensions stopped serving free, AI Pro and Ultra users on 2026-06-18. The $0 Individual plan carries the same agent models the paid ones do; no card; not offered in mainland China, Russia, Hong Kong and 21 more places; what you send may be used to train or improve models unless you turn that off; free models: `gemini-3.1-pro`, `gemini-3.8-flash`, `gemini-3.7-flash`, `gemini-3.6-flash`, `claude-opus-4.6`, `claude-sonnet-4.6`, `gpt-oss-120b`
- [Freebuff](https://mvalentsev.github.io/awesome-free-ai-coding/providers/freebuff/): Ad-funded coding agent — CLI, desktop, web, cloud and chat — with no API key and no card, on a daily budget of free model hours; no card; what you send may be used to train or improve models; free models: `glm-5.3-flash`, `deepseek-v4.1-flash`, `muse-spark-1.2`
- [Cline](https://mvalentsev.github.io/awesome-free-ai-coding/providers/cline/): Open-source coding agent for VS Code, JetBrains and the terminal; signing in to its own Cline provider unlocks a rotating set of free models, each with a daily allowance, beside pay-as-you-go credits, the $9.99 ClinePass plan and BYOK; no card; what you send may be used to train or improve models; free models: `muse-spark-1.3-contributor`; provisional since 2026-09-14
- [Crush + Charm Hyper](https://mvalentsev.github.io/awesome-free-ai-coding/providers/charm-hyper/): Charm's Crush terminal agent with Hyper, its official hosted model gateway; the free plan includes monthly Hypercredits, zero data retention; no card; what you send is not used to train models
- [Autohand Code](https://mvalentsev.github.io/awesome-free-ai-coding/providers/autohand-code/): Open-source coding agent for the terminal, VS Code, JetBrains and the desktop whose Free plan runs Autohand's own latency-first coding model with no card, its messages counted per five hours and per week; no card; free models: `fantail`; provisional since 2026-09-27
- [OpenAI Codex CLI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/openai-codex-cli/): Open-source coding CLI, free by signing in with a $0 ChatGPT Free account; local coding tasks included on all plans; no card; not offered in mainland China, Russia, Hong Kong and 24 more places
- [CodeGPT](https://mvalentsev.github.io/awesome-free-ai-coding/providers/codegpt/): VS Code / JetBrains coding agent whose $0 plan includes model usage rather than only BYOK — a small daily allowance on its own Economy models, plus BYOK across 15+ providers and local models (Ollama, LM Studio) beside it; no card

## LLM APIs with free tier

- [Google AI Studio (Gemini API)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/google-ai-studio/): Free tier on the Gemini API, priced model by model rather than as one account quota; no card; not offered in mainland China, Russia, Hong Kong and 13 more places; what you send may be used to train or improve models; key from https://aistudio.google.com/apikey; OpenAI-compatible at https://generativelanguage.googleapis.com/v1beta/openai/; free models: `gemini-3.8-flash`, `gemini-3.7-flash`, `gemini-3.6-flash`, `gemini-3.5-flash`, `gemini-3.5-flash-lite`, `gemini-3.1-flash-lite`, `gemini-3-flash`, `gemma-4`
- [Groq](https://mvalentsev.github.io/awesome-free-ai-coding/providers/groq-free/): Fast inference against a free plan Groq publishes as a per-model rate table; no card; what you send is not used to train models; key from https://console.groq.com/keys; OpenAI-compatible at https://api.groq.com/openai/v1; free models: `gpt-oss-120b`, `gpt-oss-20b`, `qwen3.8-27b`
- [NVIDIA NIM (build.nvidia.com)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/nvidia-nim/): Free endpoints for the models build.nvidia.com marks "Free Endpoint", open-weight models from several labs among them, called with a free NVIDIA Developer Program key at integrate.api.nvidia.com/v1 (OpenAI-compatible); no card; not offered in Russia, Pakistan, Bangladesh and 11 more places; what you send may be used to train or improve models; key from https://build.nvidia.com; OpenAI-compatible at https://integrate.api.nvidia.com/v1; free models: `kimi-k3`, `nemotron-3-ultra`, `nemotron-3-super`, `laguna-xs-2.1`, `muse-glimmer-30b`, `nemotron-3.5-lightning`, `gemma-4-31b`, `gpt-oss-20b`, `mistral-nemotron`, `nemotron-3-nano-omni`, `diffusiongemma`, `llama-3.2-90b-vision`, `llama-3.2-11b-vision`
- [Z.ai (Zhipu GLM)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/zai-glm/): GLM Flash models free on the API, vision included (OpenAI-compatible at api.z.ai/api/paas/v4); no card; what you send is not used to train models; key from https://z.ai/manage-apikey/apikey-list; OpenAI-compatible at https://api.z.ai/api/paas/v4; Anthropic Messages at https://api.z.ai/api/anthropic; free models: `glm-4.7-flash`, `glm-4.5-flash`, `glm-4.6v-flash`
- [Cloudflare Workers AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/cloudflare-workers-ai/): 10k neurons/day free; no card; what you send is not used to train models; key from https://dash.cloudflare.com/profile/api-tokens; OpenAI-compatible at https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/v1
- [LLM7.io](https://mvalentsev.github.io/awesome-free-ai-coding/providers/llm7/): OpenAI-compatible API with an anonymous tier — no account, no key — of 500,000 tokens a day on its turbo models; a free token doubles it; no card; no key; OpenAI-compatible at https://api.llm7.io/v1; provisional since 2026-09-16
- [LLM Tech](https://mvalentsev.github.io/awesome-free-ai-coding/providers/llmtech/): EU provider of one model whose quickstart prints a shared trial key for anyone: 2M tokens a day per address and 4 concurrent requests, tool calls included, no account; no card; what you send is not used to train models; no account: the vendor prints a key for anyone at https://llmtech.eu/docs/, `lt-trial-ba1ef28c6d32ed6980678d8d`; OpenAI-compatible at https://api.llmtech.eu/v1; free models: `qwen3.8-27b`; provisional since 2026-09-18
- [Hetzner Inference API](https://mvalentsev.github.io/awesome-free-ai-coding/providers/hetzner-inference/): OpenAI-compatible API on Hetzner's own EU hardware, free for as long as the experiment runs; no card; what you send is not used to train models; key from https://experiments.hetzner.com/inference; OpenAI-compatible at https://inference.hetzner.com/api/v1; free models: `qwen3.6`, `qwen3.8-27b`
- [Alibaba Cloud Model Studio (DashScope, international)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/alibaba-model-studio/): A million free tokens on each of its chat models in the Singapore region — Qwen, DeepSeek and GLM among them — valid 90 days from activation or the model's release; OpenAI-compatible; no card; not offered in mainland China; what you send is not used to train models; key from https://modelstudio.console.alibabacloud.com; OpenAI-compatible at https://dashscope-intl.aliyuncs.com/compatible-mode/v1; free models: `qwen3.8-max`, `qwen3.8-flash`, `qwen3.8-2.4t-a95b`, `qwen3.8-27b`, `deepseek-v4-pro`, `deepseek-v4-flash`, `qwen3-coder-plus`, `qwen3-coder-next`, `qwen3.7-max`, `qwen3.7-plus`, `qwen3.7-flash`, `qwen3.6-max`, `qwen3.6-plus`, `qwen3.6-flash`, `qwen3.6-35b-a3b`, `qwen3.6-27b`, `qwen3.5-plus`, `qwen3.5-flash`, `qwen3.5-397b-a17b`, `qwen3.5-122b-a10b`, `qwen3.5-35b-a3b`, `qwen3.5-27b`, `qwen3-max`, `qwen3-coder-flash`, `qwen3-coder-480b-a35b`, `qwen3-coder-30b-a3b`, `qwen3-next-80b-a3b-thinking`, `qwen3-next-80b-a3b-instruct`, `qwen3-235b-a22b-thinking-2507`, `qwen3-235b-a22b-instruct-2507`, `qwen3-235b-a22b`, `qwen3-30b-a3b-thinking-2507`, `qwen3-30b-a3b-instruct-2507`, `qwen3-30b-a3b`, `qwen3-32b`, `qwen3-14b`, `qwen3-8b`, `qwen-max`, `qwen-plus`, `qwen-flash`, `qwen-turbo`, `qwq-plus`, `qvq-max`, `qwen3-vl-plus`, `qwen3-vl-flash`, `qwen3-vl-235b-a22b-thinking`, `qwen3-vl-235b-a22b-instruct`, `qwen3-vl-32b-thinking`, `qwen3-vl-32b-instruct`, `qwen3-vl-30b-a3b-thinking`, `qwen3-vl-30b-a3b-instruct`, `qwen3-vl-8b-thinking`, `qwen3-vl-8b-instruct`, `qwen-vl-max`, `qwen-vl-plus`, `deepseek-v3.2`, `glm-5.1`
- [Mistral Studio](https://mvalentsev.github.io/awesome-free-ai-coding/providers/mistral-free/): Mistral's Free plan — API keys with $10 a month of included usage, shared by the API, Studio and the Vibe coding CLI, no card; no card; what you send may be used to train or improve models unless you turn that off; key from https://console.mistral.ai/api-keys; OpenAI-compatible at https://api.mistral.ai/v1
- [VLM Run Gateway](https://mvalentsev.github.io/awesome-free-ai-coding/providers/vlm-run-gateway/): OpenAI-compatible gateway for vision and language models whose models on VLM Run's own GPUs answer anonymous callers — no signup, no key — at 100 requests a day per IP, in alpha; no card; no key; OpenAI-compatible at https://gateway.vlm.run/v1/openai; free models: `qwen3.8-27b`; provisional since 2026-09-17
- [Cohere (trial keys)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/cohere/): Cohere's Command models via free trial API keys that never expire, plus a 30B/3B Apache-2.0 coding model Cohere prices at zero on every key type; no card; not offered in mainland China, Russia, Hong Kong and 5 more places; what you send may be used to train or improve models; key from https://dashboard.cohere.com/api-keys; OpenAI-compatible at https://api.cohere.com/compatibility/v1; free models: `command-a-plus`, `command-a-reasoning`, `north-mini-code`, `command-a`, `command-a-vision`, `command-r-plus`, `command-r`, `command-r7b`
- [SEA-LION (AI Singapore)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/sea-lion/): AI Singapore's open Southeast-Asian model family behind a first-party OpenAI-compatible API — the vendor hosting its own weights rather than a gateway reselling somebody else's; no card; key from https://playground.sea-lion.ai/key-manager; OpenAI-compatible at https://api.sea-lion.ai/v1
- [OpenTyphoon (SCB 10X)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/opentyphoon/): Thai-tuned open models from SCB 10X, the venture arm of Siam Commercial Bank, behind an OpenAI-compatible API whose FAQ calls it a research showcase and free to use; no card; what you send may be used to train or improve models; key from https://playground.opentyphoon.ai/api-key; OpenAI-compatible at https://api.opentyphoon.ai/v1
- [Agnes AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/agnes-ai/): Agnes AI's own models behind an OpenAI-compatible API, with its Flash text models charged at zero today and image generation free beside them; no card; what you send may be used to train or improve models unless you turn that off; key from https://platform.agnes-ai.com; OpenAI-compatible at https://apihub.agnes-ai.com/v1; free models: `agnes-3.0-flash`, `agnes-2.5-flash`; provisional since 2026-09-14
- [FreeInference (Harvard SEAS)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/freeinference/): Harvard SEAS's MadSys Lab serving open models free to every account behind both an OpenAI-shaped and an Anthropic-shaped endpoint, with a documented Claude Code setup; no card; key from https://freeinference.org; OpenAI-compatible at https://freeinference.org/v1; Anthropic Messages at https://freeinference.org/anthropic; free models: `deepseek-v4-flash`, `glm-5.1`, `glm-5.3-flash`, `minimax-m3`, `qwen3.6`, `minimax-m2.5`, `diffusiongemma`
- [Ollama Cloud](https://mvalentsev.github.io/awesome-free-ai-coding/providers/ollama-cloud/): Cloud-hosted open models on a $0 plan that grants starter usage credits for a starter subset of the catalog; no card; what you send is not used to train models; key from https://ollama.com/settings/keys; OpenAI-compatible at https://ollama.com/v1
- [OVHcloud AI Endpoints](https://mvalentsev.github.io/awesome-free-ai-coding/providers/ovh-ai-endpoints/): EU-hosted serverless open-model API whose anonymous lane needs no signup, no key and no card (OpenAI-compatible), at two requests a minute per model shared by every anonymous caller; no card; what you send is not used to train models; no key; OpenAI-compatible at https://oai.endpoints.kepler.ai.cloud.ovh.net/v1; free models: `gpt-oss-120b`, `qwen3.6`, `qwen3.8-27b`, `qwen3-coder`
- [Poolside Platform](https://mvalentsev.github.io/awesome-free-ai-coding/providers/poolside/): Free self-serve developer access to the Laguna coding models, direct from the vendor whose models this list already carries second-hand through OpenRouter and Kilo Gateway; no card; what you send may be used to train or improve models unless you turn that off; key from https://platform.poolside.ai; OpenAI-compatible at https://inference.poolside.ai/v1
- [Sail Research](https://mvalentsev.github.io/awesome-free-ai-coding/providers/sail-research/): Open-weight models for long-running agents — Kimi K3, GLM-5.3, DeepSeek V4 Pro — behind OpenAI- and Anthropic-compatible APIs, with $5 of free credit every month once a payment method is on the account; card required; not offered in Russia, Iran, Syria and 2 more places; what you send is not used to train models; key from https://app.sailresearch.com; OpenAI-compatible at https://api.sailresearch.com/v1; provisional since 2026-09-21
- [uncloseai (unturf)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/uncloseai/): Keyless OpenAI-compatible chat endpoint — no signup, no key, no account; no card; no key; OpenAI-compatible at https://hermes.ai.unturf.com/v1
- [Pollinations.AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/pollinations/): Legacy open text API, no signup, OpenAI-compatible (POST text.pollinations.ai/openai), on one model; no card; no key; OpenAI-compatible at https://text.pollinations.ai/openai; free models: `gpt-oss-20b`
- [IBM watsonx.ai (Lite plan)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/ibm-watsonx-ai/): IBM's watsonx.ai Runtime on its Lite plan — 300,000 tokens a month of foundation-model inference (Granite, Llama, Mistral and other hosted models) on IBM Cloud, a plan IBM's own docs call free and never bill; card required; not offered in Vietnam, Türkiye, Taiwan and 24 more places; what you send is not used to train models; key from https://cloud.ibm.com/iam/apikeys; API at https://us-south.ml.cloud.ibm.com/ml/v1
- [Arli AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/arli-ai/): OpenAI-compatible inference on open models and their fine-tunes, whose Free plan tries each model five times every two days at 12K tokens of context, one request at a time; no card; what you send is not used to train models; key from https://www.arliai.com/account; OpenAI-compatible at https://api.arliai.com/v1; provisional since 2026-09-17
- [Bytez](https://mvalentsev.github.io/awesome-free-ai-coding/providers/bytez/): Serverless API over open models, whose Free plan grants $1 of credit every four weeks for open models of up to 7B parameters, one request at a time, with no billing; no card; key from https://bytez.com/api; OpenAI-compatible at https://api.bytez.com/models/v2/openai/v1; provisional since 2026-09-17
- [Mixlayer](https://mvalentsev.github.io/awesome-free-ai-coding/providers/mixlayer/): Serverless open models priced per token, one of them at $0 and callable without prepaid credit; no card; what you send is not used to train models; key from https://console.mixlayer.com/app/api-keys; OpenAI-compatible at https://models.mixlayer.ai/v1; free models: `qwen3.5-4b`; provisional since 2026-09-17
- [Yolo-Auto](https://mvalentsev.github.io/awesome-free-ai-coding/providers/yolo-auto/): One open-weight Qwen model served on the vendor's own flat-rate API for coding agents; the free plan is 15 requests a week with no card; no card; what you send is not used to train models; key from https://yolo-auto.com/app; OpenAI-compatible at https://yolo-auto.com/v1; free models: `qwen3.8-flash`; provisional since 2026-09-14
- [ModelScope API-Inference (Alibaba)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/modelscope/): Alibaba's model community serves 35 models free — DeepSeek V4 Pro, GLM-5.2, MiniMax M3 and Qwen3.8 among them — for 250 魔粒 a day at 0.5 to 2 a call, after Alibaba Cloud real-name verification; no card; offered only in mainland China; key from https://modelscope.cn/my/myaccesstoken; OpenAI-compatible at https://api-inference.modelscope.cn/v1; Anthropic Messages at https://api-inference.modelscope.cn; provisional since 2026-09-19
- [SenseNova (SenseTime 商汤)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/sensenova/): SenseTime's own SenseNova models behind an OpenAI-compatible url, free for everyone while the token plan is in public beta; no card; offered only in mainland China; key from https://platform.sensenova.cn; OpenAI-compatible at https://token.sensenova.cn/v1
- [Tencent Cloud TokenHub](https://mvalentsev.github.io/awesome-free-ai-coding/providers/tencent-tokenhub/): Tencent Cloud's model platform, with a one-time grant of a million tokens on each of its language models, valid a year and no card, on an account that has passed Tencent Cloud real-name verification; no card; offered only in mainland China, Hong Kong, Taiwan and Macao; what you send is not used to train models; key from https://console.cloud.tencent.com/tokenhub/apikey; OpenAI-compatible at https://tokenhub.tencentmaas.com/v1; Anthropic Messages at https://tokenhub.tencentmaas.com; provisional since 2026-09-17
- [SiliconFlow (China)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/siliconflow-cn/): China's SiliconFlow prices eight small chat models at ¥0 for an account verified with Chinese, Hong Kong, Macau or Taiwan papers; no card; offered only in mainland China, Hong Kong, Taiwan and Macao; what you send is not used to train models; key from https://cloud.siliconflow.cn/account/ak; OpenAI-compatible at https://api.siliconflow.cn/v1; Anthropic Messages at https://api.siliconflow.cn; free models: `xing4.0-29b`, `qwen3-8b`, `glm-4-9b-0414`; provisional since 2026-09-19

## Trials (no card when possible)

- [GitHub Copilot Free](https://mvalentsev.github.io/awesome-free-ai-coding/providers/github-copilot-free/): Free Copilot plan for individual developers in VS Code, JetBrains, Visual Studio and CLI; completions, limited chat and agent usage; no card; not offered in Russia, Iran, Belarus and 2 more places; what you send may be used to train or improve models unless you turn that off
- [Kiro](https://mvalentsev.github.io/awesome-free-ai-coding/providers/kiro/): Perpetual free tier of AWS's spec-driven agentic IDE (successor to Amazon Q Developer) with Claude and open-weight models; no card; what you send may be used to train or improve models unless you turn that off; free models: `claude-sonnet-4.5`, `claude-sonnet-4`, `glm-5`, `minimax-m2.5`, `deepseek-v3.2`, `qwen3-coder-next`, `minimax-m2.1`
- [Google Jules](https://mvalentsev.github.io/awesome-free-ai-coding/providers/google-jules/): Free tier of Google's async cloud coding agent, on Gemini models; connects to GitHub repos and works autonomously; no card; what you send may be used to train or improve models unless you turn that off; free models: `gemini-2.5`
- [Cursor (Hobby)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/cursor-hobby/): Permanent free Hobby plan of the Cursor AI IDE — limited Agent requests, Cursor's own model and Tab completions, no credit card; no card; what you send may be used to train or improve models unless you turn that off; free models: `composer`
- [Gemini Enterprise Agent Platform express mode (formerly Vertex AI)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/vertex-ai-express/): Google Cloud's express mode: an API key and 90 days of Gemini models within the free tier's quotas, with no billing information, for a new Google Cloud user on a @gmail.com account; no card; what you send is not used to train models; key from https://console.cloud.google.com/expressmode; API at https://aiplatform.googleapis.com/v1; free models: `gemini-3.1-pro`, `gemini-3-flash`, `gemini-2.5-pro`, `gemini-2.5-flash`; provisional since 2026-09-23
- [Devin Desktop (formerly Windsurf)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/windsurf/): Free plan of Cognition's desktop coding agent — the IDE that shipped as Windsurf; no card; what you send may be used to train or improve models; free models: `swe-1.6`
- [Qoder](https://mvalentsev.github.io/awesome-free-ai-coding/providers/qoder/): Alibaba's agentic coding apps — the Qoder desktop agent, Qoder IDE and CLI — with a 2-week Pro trial of 300 credits on signup, then 100 credits a day to claim in the desktop app while that promotion runs; no card
- [JetBrains AI (AI Free)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/jetbrains-ai/): AI Free in JetBrains IDEs — unlimited code completion on JetBrains' own model and 3 AI Credits ($3) of cloud models every 30 days for chat and agents; no card; not offered in mainland China, Russia, Hong Kong and 25 more places; what you send may be used to train or improve models unless you turn that off; free models: `mellum`; provisional since 2026-09-14
- [Regolo AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/regolo/): EU (Italian) zero-retention inference; a month of full model access on a daily token allowance, no card; no card; what you send is not used to train models; key from https://dashboard.regolo.ai; OpenAI-compatible at https://api.regolo.ai/v1; free models: `glm-5.2`, `gpt-oss-120b`, `qwen3.8-27b`, `apertus-70b`, `gemma-4-31b`, `gpt-oss-20b`, `mistral-small-4`, `qwen3.5-9b`, `qwen3.5-122b-a10b`
- [Baidu Comate (文心快码)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/baidu-comate/): Baidu's coding assistant for VS Code, JetBrains, Visual Studio and Xcode, with its own IDE and CLI, whose free Personal Standard plan keeps code completion free and adds a one-time ¥10 voucher for agent requests; no card; provisional since 2026-09-27
- [TRAE (TraeCode)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/trae/): TRAE's AI IDE, TraeCode, whose Free plan runs Auto mode only — no model choice — on a one-time Basic usage allowance of a dollar or so that does not reset, beside 5,000 autocompletions; no card; not offered in mainland China, Russia, Canada and 11 more places; what you send may be used to train or improve models unless you turn that off
- [Upstage (Solar API)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/upstage/): Upstage Solar LLM API; $10 free credit on signup, no card; no card; what you send is not used to train models; key from https://console.upstage.ai/api-keys; OpenAI-compatible at https://api.upstage.ai/v1
- [Zed](https://mvalentsev.github.io/awesome-free-ai-coding/providers/zed/): Open-source code editor with a hosted AI agent: a 14-day Pro trial with $5 of model usage and unlimited edit predictions, no card, then 2,000 accepted edit predictions on the $0 Personal plan; no card; what you send is not used to train models; free models: `gpt-6-luna`; provisional since 2026-09-16
- [AtomCode](https://mvalentsev.github.io/awesome-free-ai-coding/providers/atomcode/): Open-source terminal coding agent from AtomGit, the code host run by CSDN and the Open Atom Foundation, whose sign-in claims a free 30-day coding plan, its quota counted in five-hour windows and its claims capped at a thousand a day; no card; free models: `deepseek-v4-flash`; provisional since 2026-09-27
- [CodeBuddy (Tencent)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/codebuddy/): Tencent's VS Code / JetBrains / CLI coding agent whose Free plan carries 100 credits a month, a daily activity bonus of 30 and 250 welcome credits, with every model open while the promotion runs — its site and the agent's own endpoint unreachable from the US, India and Russia; no card; not offered in the United States, India and Russia
- [Qoder CN (formerly Lingma)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/qoder-cn/): Alibaba Cloud's agentic coding apps for China — Qoder CN IDE, JetBrains plugin and CLI, formerly Tongyi Lingma — with a two-week trial of 300 credits, then 100 credits a day to claim while that promotion runs, and limited completions on a basic model after; no card; provisional since 2026-09-27
- [ZCode (Z.ai)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/zcode/): Z.ai's desktop coding agent, free as an app, with five days of its GLM models for a new user and no card; no card; what you send is not used to train models; free models: `glm-5.3`, `glm-5-turbo`; provisional since 2026-09-26
- [Qodo](https://mvalentsev.github.io/awesome-free-ai-coding/providers/qodo/): Agentic PR code review plus Git and IDE integrations on a 14-day Pro Team trial with unlimited reviews and credits and no card; no card
- [The Grid](https://mvalentsev.github.io/awesome-free-ai-coding/providers/the-grid/): OpenAI- and Anthropic-compatible inference market that sells quality tiers rather than model names — Agent Max was served by Claude Opus 5 in the 30 days to 2026-09-03 — with a $25 signup credit, for a limited time; no card; what you send is not used to train models; key from https://app.thegrid.ai/profile; OpenAI-compatible at https://api.thegrid.ai/v1; Anthropic Messages at https://messages-beta.api.thegrid.ai; provisional since 2026-09-17
- [Dahl Inference](https://mvalentsev.github.io/awesome-free-ai-coding/providers/dahl-inference/): An OpenAI-compatible gateway to open models served by the Gonka decentralized GPU network — GLM-5.3-Flash, DeepSeek V4 Flash and MiniMax M2.7 — whose sign-up asks for a username and nothing else and puts 100 million tokens in the account; no card; key from https://inference.dahl.global/account; OpenAI-compatible at https://inference.dahl.global/v1; provisional since 2026-09-21
- [Sarvam AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/sarvam/): India's Sarvam AI credits every new account ₹100 that never expire, spendable on any of its APIs — including its own Sarvam-105B chat model on an OpenAI-shaped endpoint; no card; what you send may be used to train or improve models unless you turn that off; key from https://dashboard.sarvam.ai; OpenAI-compatible at https://api.sarvam.ai/v1
- [Inception Labs (Mercury)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/inception-labs/): A signup grant on the Mercury diffusion models — Mercury 2.5 and Mercury 2 for chat, Mercury Edit 2 for fill-in-the-middle and code edits; the last is the reason this row is here, since an FIM endpoint is what an IDE completion plugin actually calls; no card; what you send may be used to train or improve models unless you turn that off; key from https://platform.inceptionlabs.ai; OpenAI-compatible at https://api.inceptionlabs.ai/v1
- [Blue Claw Network](https://mvalentsev.github.io/awesome-free-ai-coding/providers/blue-claw/): OpenAI-compatible endpoint that routes calls to a network of independent GPU operators running open models; every new account starts with a $5 welcome credit, and no card is asked to start; no card; key from https://portal.blueclaw.network/; OpenAI-compatible at https://openai.blueclaw.network/v1; provisional since 2026-09-17
- [HPC-AI Model APIs](https://mvalentsev.github.io/awesome-free-ai-coding/providers/hpc-ai/): OpenAI-compatible APIs over 24 models, GLM 5.3 Flash, Kimi K3 and MiniMax M3 among them, with $2 of free credit for every user — $4 with the vendor's invite code — at 5 requests a minute until a first deposit; no card; not offered in mainland China; what you send is not used to train models; key from https://www.hpc-ai.com/models-console/api-key; OpenAI-compatible at https://api.hpc-ai.com/inference/v1; provisional since 2026-09-17
- [Fireworks AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/fireworks-ai/): Serverless inference on open-weight models — Kimi K3, GLM 5.3, DeepSeek V4.1 Flash among them — with a one-time $1 of credit that is spent without a card, at 10 requests a minute until a payment method is added; no card; what you send is not used to train models; key from https://app.fireworks.ai/settings/users/api-keys; OpenAI-compatible at https://api.fireworks.ai/inference/v1; Anthropic Messages at https://api.fireworks.ai/inference; provisional since 2026-09-17
- [RouterPlex](https://mvalentsev.github.io/awesome-free-ai-coding/providers/routerplex/): A one-time $1 of free credit on a prepaid reseller that bills its catalog at vendor list prices with 0% markup, no card — one key on both wires, OpenAI-compatible Chat Completions and an Anthropic Messages base Claude Code takes as it is; no card; what you send is not used to train models; key from https://routerplex.com/sign-up; OpenAI-compatible at https://api.routerplex.com/v1; Anthropic Messages at https://api.routerplex.com
- [abliteration.ai](https://mvalentsev.github.io/awesome-free-ai-coding/providers/abliteration-ai/): OpenAI- and Anthropic-compatible API for three uncensored reasoning models, the large one derived from GLM-5.3, that opens with a one-credit free preview and no card; no card; what you send is not used to train models; key from https://abliteration.ai/console; OpenAI-compatible at https://api.abliteration.ai/v1; Anthropic Messages at https://api.abliteration.ai; provisional since 2026-09-17
- [Impossibl](https://mvalentsev.github.io/awesome-free-ai-coding/providers/impossibl/): Prepaid gateway at provider list prices, Claude, GPT, Gemini, DeepSeek and GLM among its models, whose keyless sign-up funds an account with $0.05 and adds $1 once a person claims it by email — no card; no card; key from https://impossibl.com/dashboard; OpenAI-compatible at https://api.impossibl.com/v1; Anthropic Messages at https://api.impossibl.com; provisional since 2026-09-17
- [Standard Compute](https://mvalentsev.github.io/awesome-free-ai-coding/providers/standardcompute/): A one-time $0.25 of smart-routed compute on a flat-rate agent gateway, no card — a real key on both wires, OpenAI-compatible Chat Completions and an Anthropic Messages base Claude Code takes as it is; no card; what you send is not used to train models; key from https://standardcompute.com/signup; OpenAI-compatible at https://api.stdcmpt.com/v1; Anthropic Messages at https://api.stdcmpt.com

## Aggregators (one key, many providers)

- [OpenRouter (free models)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/openrouter-free/): One API key for a rotating set of :free model variants, open-weight and stealth models among them; no card; what you send may be used to train or improve models unless you turn that off; key from https://openrouter.ai/settings/keys; OpenAI-compatible at https://openrouter.ai/api/v1; Anthropic Messages at https://openrouter.ai/api; free models: `nemotron-3-ultra`, `gemma-4-31b`, `gemma-4-26b-a4b`, `nemotron-3-super`, `north-mini-code`, `laguna-s-2.1`, `laguna-xs-2.1`, `nemotron-3.5-lightning`, `inkling`, `inkling-small`, `dots-3-note`, `ling-3.0-flash-fin`, `ling-3.0-flash-sante`, `nemotron-3-nano-omni`
- [Requesty](https://mvalentsev.github.io/awesome-free-ai-coding/providers/requesty/): OpenAI-compatible router over a 690+ model catalog with routing, caching and fallbacks; twelve rows in it are priced 0 and the free plan is the same gateway restricted to those; no card; what you send may be used to train or improve models; key from https://app.requesty.ai/api-keys; OpenAI-compatible at https://router.requesty.ai/v1; Anthropic Messages at https://router.requesty.ai; free models: `nemotron-3-ultra`, `nemotron-3-super`, `gemma-4-31b`, `ling-3.0-tiny`, `muse-glimmer-30b`, `nemotron-3.5-lightning`, `leanstral-1.5`, `nemotron-3-nano-30b`, `nemotron-3-nano-omni`
- [AIHubMix (free models)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/aihubmix/): One OpenAI-compatible gateway over 800+ models, dozens of which the platform prices at 0 and subsidises itself, with an Anthropic-format /v1/messages too, so a free id can back Claude Code; no card; key from https://aihubmix.com/token; OpenAI-compatible at https://aihubmix.com/v1; Anthropic Messages at https://aihubmix.com; free models: `glm-5.3`, `glm-5.3-flash`, `kimi-k3`, `glm-5.2`, `glm-5.1`, `glm-5`, `glm-5-turbo`, `mimo-v2.5-pro`, `mimo-v2.5`, `north-mini-code`, `nemotron-3-ultra`, `nemotron-3-super`, `minimax-m3`, `minimax-m2.7`, `hy3`, `laguna-s-2.1`, `laguna-xs-2.1`, `kimi-for-coding`, `ling-3.0-flash`, `ling-3.0-tiny`, `glm-4.7`, `glm-4.7-flash`, `glm-4.6`, `minimax-m2.5`, `minimax-m2.1`, `minimax-m2`, `mimo-v2-pro`, `mimo-v2-flash`, `mimo-v2-omni`, `dots-3-note`, `nemotron-3.5-lightning`, `nemotron-3-nano-30b`, `nemotron-3-nano-omni`, `nemotron-nano-12b-v2-vl`, `nemotron-nano-9b-v2`
- [Routeway](https://mvalentsev.github.io/awesome-free-ai-coding/providers/routeway/): OpenAI-compatible gateway whose :free lane rotates — three zero-priced chat models on 2026-09-23 — beside 269 metered rows in the same catalog; no card; key from https://routeway.ai/dashboard/keys; OpenAI-compatible at https://api.routeway.ai/v1; free models: `deepseek-v4-flash`, `minimax-m2.7`, `muse-glimmer-30b`
- [LLMTR](https://mvalentsev.github.io/awesome-free-ai-coding/providers/llmtr/): Turkish OpenAI-compatible gateway whose free rows answer on a zero balance — ten zero-priced chat ids on 2026-09-27; no card; what you send may be used to train or improve models; key from https://llmtr.com/dashboard/api-keys; OpenAI-compatible at https://llmtr.com/v1; free models: `nemotron-3-ultra`, `nemotron-3-super`, `laguna-xs-2.1`, `ling-3.0-flash-fin`
- [Vercel AI Gateway](https://mvalentsev.github.io/awesome-free-ai-coding/providers/vercel-ai-gateway/): One OpenAI-compatible endpoint for 360+ models, with $5 of gateway credits every month once the team has a payment method on file, and three language models priced at zero that never touch the credit; card required; what you send may be used to train or improve models unless you turn that off; key from https://vercel.com/dashboard/ai-gateway/api-keys; OpenAI-compatible at https://ai-gateway.vercel.sh/v1; Anthropic Messages at https://ai-gateway.vercel.sh; free models: `laguna-s-2.1`, `ling-3.0-flash-sante`
- [BazaarLink](https://mvalentsev.github.io/awesome-free-ai-coding/providers/bazaarlink/): OpenAI-compatible gateway to a 128-id catalog whose free page counts two models on 2026-09-27, beside the auto:free router; no card; key from https://bazaarlink.ai/keys; OpenAI-compatible at https://api.bazaarlink.ai/v1; free models: `qwen3.7-flash`
- [Nous Portal (Hermes Agent)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/nous-portal/): Nous Research's inference portal behind its Hermes Agent: a $0 Free plan limited to the models it prices at zero — seven on 2026-09-18 — on an OpenAI-compatible API; no card; what you send may be used to train or improve models unless you turn that off; key from https://portal.nousresearch.com; OpenAI-compatible at https://inference-api.nousresearch.com/v1; free models: `step-3.7-flash`, `laguna-s-2.1`; provisional since 2026-09-16
- [Token Harbor](https://mvalentsev.github.io/awesome-free-ai-coding/providers/token-harbor/): OpenAI- and Anthropic-compatible gateway with a standing $0 plan: a rotating lineup of :free ids on a value-based allowance per rolling 7-day period, no card; no card; what you send may be used to train or improve models; key from https://tokenharbor.ai/dashboard/api-keys; OpenAI-compatible at https://tokenharbor.ai/v1; Anthropic Messages at https://tokenharbor.ai; free models: `deepseek-v4.1-flash`; provisional since 2026-09-16
- [TokenRouter (PaleBlueDot)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/tokenrouter/): One zero-priced id, in the default group, inside a 140-row catalog that meters the rest; no card; not offered in Russia, Iran, Belarus and 3 more places; key from https://www.tokenrouter.com/console/token; OpenAI-compatible at https://api.tokenrouter.com/v1; free models: `nemotron-3-nano-omni`
- [Hugging Face Inference Providers](https://mvalentsev.github.io/awesome-free-ai-coding/providers/huggingface-inference/): Routed access to 200+ models across providers (Groq, Cerebras, Together, etc.) with a free HF account; no card; key from https://huggingface.co/settings/tokens; OpenAI-compatible at https://router.huggingface.co/v1
- [Opper](https://mvalentsev.github.io/awesome-free-ai-coding/providers/opper/): EU-hosted gateway over 700+ models whose free models answer an account with no card on file; every other model needs a card and credits; no card; what you send is not used to train models; key from https://platform.opper.ai/settings/api-keys; OpenAI-compatible at https://api.opper.ai/v3/compat; Anthropic Messages at https://api.opper.ai/v3/compat; free models: `gemma-4-31b`; provisional since 2026-09-17
- [MegaNova](https://mvalentsev.github.io/awesome-free-ai-coding/providers/meganova/): OpenAI-compatible gateway whose no-card Tier 1 account gets 50 free requests a day on each of its free rows — a chat model and the house Manta routers among them — 550 a day in all; no card; not offered in Hong Kong, Venezuela, Azerbaijan and 40 more places; key from https://www.meganova.ai/api-keys; OpenAI-compatible at https://api.meganova.ai/v1; free models: `mistral-small-3.2`
- [Moark (Gitee AI)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/moark/): Gitee's model platform, formerly Gitee AI — 200+ open models behind OpenAI- and Anthropic-compatible APIs — whose free experience token gives every user 100 calls a day across its featured models, with nothing bought; no card; what you send is not used to train models; key from https://moark.com/dashboard/tokens; OpenAI-compatible at https://api.moark.com/v1; Anthropic Messages at https://moark.com/anthropic; provisional since 2026-09-17
- [Experiential Labs](https://mvalentsev.github.io/awesome-free-ai-coding/providers/experiential-labs/): An open-source AI gateway backed by Y Combinator — every hosted provider behind one OpenAI-compatible key at the providers' list prices — whose free plan carries 500 credits ($5) a month after a one-time $1 card check; card required; key from https://platform.experientiallabs.ai; OpenAI-compatible at https://api.experientiallabs.ai/v1; provisional since 2026-09-21

## Free models by name

A model gets a page of its own once two rows or more serve it free, or it measures notable, strong or frontier: every row that serves it, the limits in the vendor's words and the ids to call. Every other model is named on the one row above that serves it.

- [nemotron-3-ultra](https://mvalentsev.github.io/awesome-free-ai-coding/models/nemotron-3-ultra/): notable; opencode, OpenRouter (free models), Kilo Code, Requesty, NVIDIA NIM (build.nvidia.com), AIHubMix (free models), LLMTR
- [qwen3.8-27b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.8-27b/): strong; Groq, LLM Tech, Hetzner Inference API, Alibaba Cloud Model Studio (DashScope, international), Regolo AI, VLM Run Gateway, OVHcloud AI Endpoints
- [nemotron-3-nano-omni](https://mvalentsev.github.io/awesome-free-ai-coding/models/nemotron-3-nano-omni/): OpenRouter (free models), Kilo Code, Requesty, NVIDIA NIM (build.nvidia.com), AIHubMix (free models), TokenRouter (PaleBlueDot)
- [nemotron-3-super](https://mvalentsev.github.io/awesome-free-ai-coding/models/nemotron-3-super/): OpenRouter (free models), Kilo Code, Requesty, NVIDIA NIM (build.nvidia.com), AIHubMix (free models), LLMTR
- [nemotron-3.5-lightning](https://mvalentsev.github.io/awesome-free-ai-coding/models/nemotron-3.5-lightning/): opencode, OpenRouter (free models), Kilo Code, Requesty, NVIDIA NIM (build.nvidia.com), AIHubMix (free models)
- [gemma-4-31b](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemma-4-31b/): notable; OpenRouter (free models), Requesty, NVIDIA NIM (build.nvidia.com), Regolo AI, Opper
- [laguna-s-2.1](https://mvalentsev.github.io/awesome-free-ai-coding/models/laguna-s-2.1/): OpenRouter (free models), Kilo Code, AIHubMix (free models), Vercel AI Gateway (card required), Nous Portal (Hermes Agent)
- [laguna-xs-2.1](https://mvalentsev.github.io/awesome-free-ai-coding/models/laguna-xs-2.1/): OpenRouter (free models), Kilo Code, NVIDIA NIM (build.nvidia.com), AIHubMix (free models), LLMTR
- [deepseek-v4-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/deepseek-v4-flash/): strong; Routeway, Alibaba Cloud Model Studio (DashScope, international), AtomCode, FreeInference (Harvard SEAS)
- [gpt-oss-120b](https://mvalentsev.github.io/awesome-free-ai-coding/models/gpt-oss-120b/): Groq, Google Antigravity, Regolo AI, OVHcloud AI Endpoints
- [gpt-oss-20b](https://mvalentsev.github.io/awesome-free-ai-coding/models/gpt-oss-20b/): Groq, NVIDIA NIM (build.nvidia.com), Regolo AI, Pollinations.AI
- [ling-3.0-flash-fin](https://mvalentsev.github.io/awesome-free-ai-coding/models/ling-3.0-flash-fin/): notable; opencode, OpenRouter (free models), Kilo Code, LLMTR
- [north-mini-code](https://mvalentsev.github.io/awesome-free-ai-coding/models/north-mini-code/): OpenRouter (free models), Kilo Code, AIHubMix (free models), Cohere (trial keys)
- [dots-3-note](https://mvalentsev.github.io/awesome-free-ai-coding/models/dots-3-note/): OpenRouter (free models), Kilo Code, AIHubMix (free models)
- [glm-5.1](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5.1/): notable; AIHubMix (free models), Alibaba Cloud Model Studio (DashScope, international), FreeInference (Harvard SEAS)
- [glm-5.3-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5.3-flash/): strong; Freebuff, AIHubMix (free models), FreeInference (Harvard SEAS)
- [ling-3.0-flash-sante](https://mvalentsev.github.io/awesome-free-ai-coding/models/ling-3.0-flash-sante/): OpenRouter (free models), Kilo Code, Vercel AI Gateway (card required)
- [minimax-m2.5](https://mvalentsev.github.io/awesome-free-ai-coding/models/minimax-m2.5/): notable; Kiro, AIHubMix (free models), FreeInference (Harvard SEAS)
- [muse-glimmer-30b](https://mvalentsev.github.io/awesome-free-ai-coding/models/muse-glimmer-30b/): notable; Requesty, NVIDIA NIM (build.nvidia.com), Routeway
- [qwen3.6](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.6/): Hetzner Inference API, FreeInference (Harvard SEAS), OVHcloud AI Endpoints
- [deepseek-v3.2](https://mvalentsev.github.io/awesome-free-ai-coding/models/deepseek-v3.2/): notable; Kiro, Alibaba Cloud Model Studio (DashScope, international)
- [deepseek-v4.1-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/deepseek-v4.1-flash/): strong; Freebuff, Token Harbor
- [diffusiongemma](https://mvalentsev.github.io/awesome-free-ai-coding/models/diffusiongemma/): NVIDIA NIM (build.nvidia.com), FreeInference (Harvard SEAS)
- [gemini-3-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3-flash/): notable; Google AI Studio (Gemini API), Gemini Enterprise Agent Platform express mode (formerly Vertex AI)
- [gemini-3.1-pro](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.1-pro/): notable; Google Antigravity, Gemini Enterprise Agent Platform express mode (formerly Vertex AI)
- [gemini-3.6-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.6-flash/): strong; Google AI Studio (Gemini API), Google Antigravity
- [gemini-3.7-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.7-flash/): strong; Google AI Studio (Gemini API), Google Antigravity
- [gemini-3.8-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.8-flash/): strong; Google AI Studio (Gemini API), Google Antigravity
- [glm-4.7-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-4.7-flash/): notable; AIHubMix (free models), Z.ai (Zhipu GLM)
- [glm-5](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5/): notable; Kiro, AIHubMix (free models)
- [glm-5-turbo](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5-turbo/): notable; AIHubMix (free models), ZCode (Z.ai)
- [glm-5.2](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5.2/): strong; AIHubMix (free models), Regolo AI
- [glm-5.3](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-5.3/): strong; AIHubMix (free models), ZCode (Z.ai)
- [inkling-small](https://mvalentsev.github.io/awesome-free-ai-coding/models/inkling-small/): notable; OpenRouter (free models), Kilo Code
- [kimi-k3](https://mvalentsev.github.io/awesome-free-ai-coding/models/kimi-k3/): strong; NVIDIA NIM (build.nvidia.com), AIHubMix (free models)
- [ling-3.0-tiny](https://mvalentsev.github.io/awesome-free-ai-coding/models/ling-3.0-tiny/): notable; Requesty, AIHubMix (free models)
- [mimo-v2.5](https://mvalentsev.github.io/awesome-free-ai-coding/models/mimo-v2.5/): notable; opencode, AIHubMix (free models)
- [minimax-m2.1](https://mvalentsev.github.io/awesome-free-ai-coding/models/minimax-m2.1/): notable; Kiro, AIHubMix (free models)
- [minimax-m2.7](https://mvalentsev.github.io/awesome-free-ai-coding/models/minimax-m2.7/): notable; AIHubMix (free models), Routeway
- [minimax-m3](https://mvalentsev.github.io/awesome-free-ai-coding/models/minimax-m3/): notable; AIHubMix (free models), FreeInference (Harvard SEAS)
- [muse-spark-1.3-contributor](https://mvalentsev.github.io/awesome-free-ai-coding/models/muse-spark-1.3-contributor/): strong; opencode, Cline
- [nemotron-3-nano-30b](https://mvalentsev.github.io/awesome-free-ai-coding/models/nemotron-3-nano-30b/): Requesty, AIHubMix (free models)
- [qwen3-8b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3-8b/): Alibaba Cloud Model Studio (DashScope, international), SiliconFlow (China)
- [qwen3-coder-next](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3-coder-next/): Kiro, Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.5-122b-a10b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.5-122b-a10b/): notable; Alibaba Cloud Model Studio (DashScope, international), Regolo AI
- [qwen3.7-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.7-flash/): Alibaba Cloud Model Studio (DashScope, international), BazaarLink
- [qwen3.8-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.8-flash/): strong; Alibaba Cloud Model Studio (DashScope, international), Yolo-Auto
- [step-3.7-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/step-3.7-flash/): notable; Kilo Code, Nous Portal (Hermes Agent)
- [claude-opus-4.6](https://mvalentsev.github.io/awesome-free-ai-coding/models/claude-opus-4.6/): notable; Google Antigravity
- [claude-sonnet-4](https://mvalentsev.github.io/awesome-free-ai-coding/models/claude-sonnet-4/): notable; Kiro
- [claude-sonnet-4.5](https://mvalentsev.github.io/awesome-free-ai-coding/models/claude-sonnet-4.5/): notable; Kiro
- [claude-sonnet-4.6](https://mvalentsev.github.io/awesome-free-ai-coding/models/claude-sonnet-4.6/): notable; Google Antigravity
- [deepseek-v4-pro](https://mvalentsev.github.io/awesome-free-ai-coding/models/deepseek-v4-pro/): strong; Alibaba Cloud Model Studio (DashScope, international)
- [gemini-2.5-pro](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-2.5-pro/): notable; Gemini Enterprise Agent Platform express mode (formerly Vertex AI)
- [gemini-3.1-flash-lite](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.1-flash-lite/): notable; Google AI Studio (Gemini API)
- [gemini-3.5-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.5-flash/): notable; Google AI Studio (Gemini API)
- [gemini-3.5-flash-lite](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemini-3.5-flash-lite/): notable; Google AI Studio (Gemini API)
- [gemma-4-26b-a4b](https://mvalentsev.github.io/awesome-free-ai-coding/models/gemma-4-26b-a4b/): notable; OpenRouter (free models)
- [glm-4.6](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-4.6/): notable; AIHubMix (free models)
- [glm-4.7](https://mvalentsev.github.io/awesome-free-ai-coding/models/glm-4.7/): notable; AIHubMix (free models)
- [gpt-6-luna](https://mvalentsev.github.io/awesome-free-ai-coding/models/gpt-6-luna/): strong; Zed
- [hy3](https://mvalentsev.github.io/awesome-free-ai-coding/models/hy3/): notable; AIHubMix (free models)
- [inkling](https://mvalentsev.github.io/awesome-free-ai-coding/models/inkling/): notable; OpenRouter (free models)
- [ling-3.0-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/ling-3.0-flash/): notable; AIHubMix (free models)
- [mimo-v2-flash](https://mvalentsev.github.io/awesome-free-ai-coding/models/mimo-v2-flash/): notable; AIHubMix (free models)
- [mimo-v2-omni](https://mvalentsev.github.io/awesome-free-ai-coding/models/mimo-v2-omni/): notable; AIHubMix (free models)
- [mimo-v2-pro](https://mvalentsev.github.io/awesome-free-ai-coding/models/mimo-v2-pro/): notable; AIHubMix (free models)
- [mimo-v2.5-pro](https://mvalentsev.github.io/awesome-free-ai-coding/models/mimo-v2.5-pro/): notable; AIHubMix (free models)
- [minimax-m2](https://mvalentsev.github.io/awesome-free-ai-coding/models/minimax-m2/): notable; AIHubMix (free models)
- [muse-spark-1.2](https://mvalentsev.github.io/awesome-free-ai-coding/models/muse-spark-1.2/): strong; Freebuff
- [qwen3-max](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3-max/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.5-27b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.5-27b/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.5-35b-a3b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.5-35b-a3b/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.5-397b-a17b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.5-397b-a17b/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.5-9b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.5-9b/): notable; Regolo AI
- [qwen3.6-27b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.6-27b/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.6-35b-a3b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.6-35b-a3b/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.6-max](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.6-max/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.6-plus](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.6-plus/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.7-max](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.7-max/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.7-plus](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.7-plus/): notable; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.8-2.4t-a95b](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.8-2.4t-a95b/): strong; Alibaba Cloud Model Studio (DashScope, international)
- [qwen3.8-max](https://mvalentsev.github.io/awesome-free-ai-coding/models/qwen3.8-max/): strong; Alibaba Cloud Model Studio (DashScope, international)

## Archived

- [AI21 Labs (Jamba)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/ai21-labs/): no longer listed — vendor-announced shutdown on 2026-08-09
- [Aider](https://mvalentsev.github.io/awesome-free-ai-coding/providers/aider/): no longer listed — delisted on 2026-07-19: BYOK only — the tool is free and open source but bundles no model access of its own, and its docs send readers to other vendors' free tiers, which this list carries directly
- [Amazon Q Developer](https://mvalentsev.github.io/awesome-free-ai-coding/providers/amazon-q-developer/): no longer listed — vendor-announced shutdown on 2026-05-15
- [Cerebras Inference](https://mvalentsev.github.io/awesome-free-ai-coding/providers/cerebras-free/): no longer listed — delisted on 2026-09-16: its only free offer was $5 of credit granted after a verified payment method is added, expiring in 30 days — a one-off credit behind a card, which CONTRIBUTING does not admit
- [DeepSeek Platform](https://mvalentsev.github.io/awesome-free-ai-coding/providers/deepseek/): no longer listed — delisted on 2026-07-27: the 5M-token signup grant was only ever reported by third parties, and DeepSeek's own pricing page prices every model and names no free grant
- [Easy GonkaAI API](https://mvalentsev.github.io/awesome-free-ai-coding/providers/easy-gonka-api/): no longer listed — delisted on 2026-07-19: rejected for cause — its /for-agents page carried a prompt injection aimed at AI agents (auto-signup, credential exfiltration, referral spam)
- [GitHub Models](https://mvalentsev.github.io/awesome-free-ai-coding/providers/github-models/): no longer listed — vendor-announced shutdown on 2026-06-16
- [Infomaniak AI Services](https://mvalentsev.github.io/awesome-free-ai-coding/providers/infomaniak-ai/): no longer listed — delisted on 2026-09-16: a one-month wallet of a million credits, and "A credit card is required to start using the API" — a one-off credit behind a card, which CONTRIBUTING does not admit
- [Kenari](https://mvalentsev.github.io/awesome-free-ai-coding/providers/kenari/): no longer listed — delisted on 2026-09-16: rejected for cause — the operator's own JavaScript bundle showed its capacity coming from pooled ChatGPT and Codex OAuth credentials, captcha solvers and a proxy pool that multiplies per-IP free quotas
- [LongCat API Platform](https://mvalentsev.github.io/awesome-free-ai-coding/providers/longcat/): no longer listed — delisted on 2026-08-14: the daily free quota was gone before the row was listed — the change log retired the Flash line on 2026-05-29 and switched billing on for LongCat-2.0 on 2026-06-30, and no LongCat page still published the 100K tokens a day
- [MiMo Code](https://mvalentsev.github.io/awesome-free-ai-coding/providers/mimo-code/): no longer listed — vendor-announced shutdown on 2026-07-26
- [Novita AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/novita/): no longer listed — delisted on 2026-08-14: the free lane ended without an announcement — both models the row named free were billed, and not one of the 102 prices on novita.ai/pricing was zero
- [Puter.js (free LLM API)](https://mvalentsev.github.io/awesome-free-ai-coding/providers/puter-free/): no longer listed — delisted on 2026-07-19: a browser SDK billed to each end user's Puter account, with no HTTP endpoint a coding agent could call; the OpenAI-compatible endpoint read on 2026-09-14 needs a paid plan
- [Reka AI](https://mvalentsev.github.io/awesome-free-ai-coding/providers/reka/): no longer listed — delisted on 2026-08-11: the $10 of free credits a month rested on one dated announcement post; no live Reka page corroborated a recurring grant, and the platform sells pay-as-you-go credits
- [SambaNova Cloud](https://mvalentsev.github.io/awesome-free-ai-coding/providers/sambanova-cloud/): no longer listed — delisted on 2026-09-23: new accounts get no free usage — the console's Free plan has read "Add a payment method and purchase credits to run your first requests" since 2026-08, where until 2026-06-20 it read "free API credits. No credit card required"; only the docs still describe a free tier
- [Scaleway Generative APIs](https://mvalentsev.github.io/awesome-free-ai-coding/providers/scaleway-generative/): no longer listed — delisted on 2026-09-16: 1,000,000 free tokens once per customer, and "Ordering Scaleway resources requires a valid credit card" — a one-off credit behind a card, which CONTRIBUTING does not admit

## Machine-readable

- [index.json](https://mvalentsev.github.io/awesome-free-ai-coding/index.json): every row with its connection details and probe, plus the watchlist of services considered and not listed
- [feed.xml](https://mvalentsev.github.io/awesome-free-ai-coding/feed.xml): Atom feed of every change — rows arriving, leaving and changing their free models
- [Provider pages](https://mvalentsev.github.io/awesome-free-ai-coding/providers/): one page per row, live and archived
- [Model pages](https://mvalentsev.github.io/awesome-free-ai-coding/models/): every free model on the list and the rows that serve it, and a page of its own for each model once two rows or more serve it free, or it measures notable, strong or frontier
- [Filterable table](https://mvalentsev.github.io/awesome-free-ai-coding/browse.html): the same rows filtered by category, card, key and API format
- [configs/opencode.json](https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/configs/opencode.json): opencode config with every OpenAI-compatible row wired up
- [configs/claude-code.sh](https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/configs/claude-code.sh): one shell function per gateway that serves the Anthropic Messages format, for Claude Code
- [configs/litellm.yaml](https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/configs/litellm.yaml): LiteLLM proxy config over the same rows, with one-name fallback groups (free/strong, free/nokey)
- [configs/codex/litellm.config.toml](https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/configs/codex/litellm.config.toml): Codex CLI profile over litellm.yaml — Codex speaks only the Responses API, which the proxy answers from each lane's chat completions
- [README](https://github.com/mvalentsev/awesome-free-ai-coding): the list itself, with the picks table and how it stays fresh
- [CONTRIBUTING](https://github.com/mvalentsev/awesome-free-ai-coding/blob/main/CONTRIBUTING.md): what qualifies, how rows are ranked, how the probes work