7.4 billion tokens per month. 34 free LLM providers. 635 free model endpoints. All behind one /v1 endpoint, plus any custom OpenAI-compatible endpoint. Smart routing, automatic failover, encrypted keys. Personal experimentation only.
TypeScript
Commits on main
fix(stream): commit long-reasoning streams early so idle clients stay connected (#1642) (#1645)
4836734 Fangore committed 28 days ago
feat(providers): add IteraCompute free OpenAI-compatible gateway (#1649) (#1655)
2359d30 tashdroid committed 29 days ago
feat(security): Host-header allowlist against DNS rebinding (ALLOWED_HOSTS, #1637) (#1644)
298a59c tashdroid committed 29 days ago
docs(readme): credit Alexis12119, kevin9327 and ilya-bogin-keenable (#1654)
ec6470b Tashfeen committed 29 days ago
fix(keys): read `export NAME=value` lines in an imported .env (#1569)
a2c81c5 kevin committed 29 days ago
feat(search): Keenable as a keyless web search engine (#1594)
7c9e26f Ilya Bogin committed 29 days ago
fix(search): a served search spends its key's monthly budget (#1158) (#1648)
2474853 tashdroid committed 29 days ago
feat(keys): set a per-key proxy from the Edit key dialog (#1526) (#1598)
7e9bfec Aymen Jeddou committed 29 days ago
fix(client): show an example base URL in the custom provider URL error (#1554) (#1653)
af3709d Tashfeen committed 29 days ago
fix(serpsearch): never send a key-check request (#1599) (#1635)
5f4b596 Tashfeen committed 29 days ago
fix(rerank): enforce the per-key monthly caps before dispatch (#1627)
d9cccbb tashdroid committed 1 month ago
fix(rerank): log rerank traffic to the requests table (#1621)
eb890c4 tashdroid committed 1 month ago
feat(models): serve GET /v1/models/:id (OpenAI models.retrieve) from the shared catalog builder (#1632)
d1eeab2 tashdroid committed 1 month ago
fix(router): the usable-key probes honor the per-key monthly budget (#1626)
938efc7 tashdroid committed 1 month ago
fix(moderation): enforce monthly caps and mark a served moderation as key evidence (#1625)
b06de14 tashdroid committed 1 month ago
fix(health): a served embedding/rerank/decision/media request marks its key working (#1620)
ec91238 tashdroid committed 1 month ago
feat(proxy): /v1/usage reports per-key spend with monthly caps (#1568) (#1614)
68ae4f8 tashdroid committed 1 month ago
feat(providers): add onomeo free OpenAI-compatible gateway (#1460) (#1610)
215cd91 tashdroid committed 1 month ago
feat(models): proactive per-model probing with explicit status labels (#1438) (#1492)
6c021c1 suantea (asuan_suan) committed 1 month ago
feat(cli): detect the installed Codex CLI / ChatGPT desktop app in setup notes (#1443)
79b03e9 suantea (asuan_suan) committed 1 month ago
fix(ratelimit): ModelScope 免费额度更新为魔粒计费,账号级每日上限 1800→100 (#1045)
10662df suantea (asuan_suan) committed 1 month ago
Codex/fix model discovery pricing main review (#1528)
10d8b00 Omansh Bhatnagar committed 1 month ago
fix(freeai): skip unverifiable credential health probes (#1540)
e6cee44 Vladislav committed 1 month ago