LLM7.io
conditional free · aggregator · verification: partially_verified (60%, doc_fetch, 2026-07-17)
The record
| Base URLs | https://api.llm7.io/v1 openai-compatible |
|---|---|
| API key required | yes |
| Credit card | no |
| Phone verification | unknown |
| May train on prompts | unknown |
| Quota | free_tier [rpm] “30 RPM (120 with token)” |
| Signup friction | steps unknown |
| Env var | LLM7_API_KEY |
| Notes | Catalog churn is high: the mnfst snapshot (2026-06-15) listed deepseek-r1-0528, gpt-4o-mini, gemini-2.5-flash-lite etc., but the live /v1/models endpoint (2026-07-17) returns a different 14-model list (gpt-5.x, claude-*, deepseek-v4-flash, kimi-k3, ...). Model records below reflect the live endpoint. API key is optional (env name recorded for the optional rate-limit token). Probe 2026-07-17: GET /models is anonymous but chat completions return 401 without a token — free token via token.llm7.io. |
Models (4)
| model | capabilities | context | free | limits |
|---|---|---|---|---|
| deepseek-v4-flash | chat, tool_calling, reasoning | ? | yes | 30 rpm |
| gpt-5.4-mini | chat, tool_calling, vision, reasoning | ? | yes | 30 rpm |
| gpt-oss:20b | chat, tool_calling, reasoning | ? | yes | 30 rpm |
| kimi-k3 | chat, tool_calling | ? | yes | 30 rpm |
Provenance
| source_llm7_docs | official_models | trust: high | checked 2026-07-17 |
| source_repo_mnfst | curated_repo | trust: medium_high | checked 2026-07-17 |