Chutes.ai
partially verifiedconditional free · inference_host · confidence 70% · mixed · last verified 2026-08-17
Subject record
| Base URLs | https://llm.chutes.ai/v1 openai-compatible |
|---|---|
| API key required | yes |
| Credit card | unknown |
| Phone verification | unknown |
| May train on prompts | unknown |
| Quota | free_tier [—] “Varies (community) | Subject to availability | Free (community-powered) | No hard cap” |
| Signup friction | steps unknown |
| Env var | CHUTES_API_KEY |
| Notes | TOMBSTONE / PIVOTED — DO NOT ROUTE FREE TRAFFIC HERE WITHOUT RE-VERIFICATION. Free-LLM README lists Chutes as a free community-powered provider (DeepSeek-R1, Llama 3.1 70B, Qwen 2.5 72B), but the official chutes.ai homepage and /pricing page (both fetched 2026-07-17) advertise NO free tier: pay-per-token public inference ('Pay per token. No subscription, no minimum, no markup.'), Plus $10/mo and Pro $20/mo plans with bundled daily quotas, and per-second private GPU deployments. Chutes historically gated free access behind a small deposit, so category is best-guessed conditional_free, but any current free path is unconfirmed. Decentralized (Bittensor-powered) compute: reliability and trust depend on the miner network — expect variable latency/availability vs centralized hosts. Model records intentionally skipped per curation policy for pivoted providers. Secondary base URL seen in Free-LLM code example: https://api.chutes.ai/v1. OPERATOR MANUAL CHECK 2026-07-25 (Peter): account creation + API key issuance WORK (a key was minted and fingerprinted into the local pool) — so the site is live and hands out keys, but this does NOT establish a free inference tier; the pay-per-token pivot (2026-07-17 pricing fetch) still stands. TOMBSTONE HOLDS. Because a real key now exists, this is settleable by a live probe: run `pnpm modelhubby probe --live --provider chutes` to see whether the key returns free completions or hits a paywall, then re-categorize from evidence. LIVE PROBE 2026-07-25: the key AUTHENTICATED against /models (200), so the account + key are valid and the provider is alive — status nudged to partially_verified. BUT no completion was obtained (synthetic not run), so the paywall-vs-free question is STILL open; the tombstone caveat holds until a real completion succeeds without payment. Connectivity ≠ free access. LIVE /models 2026-07-25 SETTLES IT: every model in the response carries a USD price (e.g. Qwen3-32B ~$0.104 in / $0.416 out per Mtok) — pay-per-token confirmed from the API itself, no free tier. The tombstone is now evidence-backed, not inferred. |
Models (0)
| model | capabilities | context | free | limits |
|---|
Provenance
| source_chutes_pricing | official_pricing | trust: high | checked 2026-07-17 |
| source_repo_freellm | curated_repo | trust: low | checked 2026-07-18 |