awesome-free-ai-coding · every free model · every provider · filter the list

FreeInference (Harvard SEAS) free tier

🔌 LLM APIs with free tier · no card · live — last verified by a probe on 2026-11-01 · freeinference.org · back to the whole list

What you get

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

Free models

deepseek-v4-flash, glm-5.3-flash, minimax-m3, qwen3.6

Limits, in the vendor’s words

Usage allowance: amount varies (period not published) per account

No card. This is an experimental research service with capacity-dependent limits. Prompts and responses may be logged for research; sanitized content, usage statistics and routing metrics may be published or open-sourced. Only models marked Free accept a Free key; Pro catalog rows need a Pro-enabled key.

Where it is offered

The vendor names no country it keeps the offer from (source, read 2026-09-26).

Connect

Base URL: https://freeinference.org/v1
Key: FREEINFERENCE_API_KEY — get one at https://freeinference.org
Anthropic-format base (Claude Code’s ANTHROPIC_BASE_URL): https://freeinference.org/anthropic
Callable ids: glm-5.3-flash, deepseek-v4-flash, qwen3.6-35b, minimax-m3
Note: Use a model marked Free in the documentation and returned for your key by authenticated /v1/models. The public view omits qwen3.6-35b and minimax-m3 although the docs mark both Free; confirm their availability for your key. Catalog prices are upstream accounting values, not fees charged to Free users.

Try it from your terminal with your key in FREEINFERENCE_API_KEY — it goes from your machine to the vendor and nowhere else:

curl -s https://freeinference.org/v1/chat/completions \
  -H "Authorization: Bearer $FREEINFERENCE_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"model":"glm-5.3-flash","messages":[{"role":"user","content":"2+2? MAKE NO MISTAKES."}]}'

Evidence

Probe: the page at https://doc.freeinference.org/models, anchored on free accounts can use models marked
Source: https://freeinference.org/
Source: https://doc.freeinference.org/models
Source: https://doc.freeinference.org/claude-code.html
Source: https://freeinference.org/terms
Source: https://freeinference.org/v1/models

History

Each line is a change to what this page publishes, dated the day it reached the list, in UTC.

2026-10-26 — Free models changed: dropped diffusiongemma, glm-5.1, minimax-m2.5
2026-09-24 — Free models changed: added diffusiongemma, minimax-m2.5
2026-09-05 — Added: Harvard SEAS’s MadSys Lab serving frontier open models — DeepSeek V4 Flash, GLM-5.1, GLM 5.3 Flash, MiniMax M3, Qwen3.6 35B — free to every account behind both an OpenAI-shaped and an Anthropic-shaped endpoint, with a documented Claude Code setup

Generated from registry.yaml on 2026-11-01 and re-verified twice a week; the full list, the Atom feed and the machinery are at https://github.com/mvalentsev/awesome-free-ai-coding.

Something wrong or missing? Suggest a service or a change — or read how the list works.