Frontier-class AI, open-source prices
Hosted open models from $0.14 per million tokens. No subscription, no minimum, no hidden markups — you pay for exactly the tokens you use.
Today's prices
Catalog prices, refreshed daily. Per million tokens.
| Model | Input | Output | Similar Arena Elo to | List price |
|---|---|---|---|---|
| DeepSeek V4 Flash | $0.14 | $0.28 | GPT-5.1 Arena Elo 1435.5 vs 1438.6 | ≈88% lower |
| GLM-4.5-Air | $0.20 | $1.10 | ||
| Gemini 2.5 Flash | $0.30 | $2.50 | Claude Haiku 4.5 Arena Elo 1410.2 vs 1412.2 | ≈50% lower |
| DeepSeek V4 Pro | $0.44 | $0.87 | Claude Sonnet 4.5 Arena Elo 1456.6 vs 1455.1 | ≈85% lower |
| GLM-4.6 | $0.60 | $2.20 | Claude Haiku 4.5 Arena Elo 1425.2 vs 1412.2 | ≈40% lower |
| Claude Haiku 4.5 | $0.80 | $4.00 |
Similarity means one thing here: LMArena Text Arena Elo (style control), our models and peers from the same snapshot dated 2026-07-21. It is not a claim of parity on coding, reasoning, latency, safety or features.
Peer prices are the vendors' published per-token API list prices, read 2026-07-31; sources in docs. This compares list prices only — actual task cost also depends on tokenizer, prompt and output length, and caching.
Badges use whichever of input/output is least favourable to us, rounded against us, and stop rendering once the peer prices are more than 30 days old.
Third-party model names are trademarks of their respective owners. Tokligence is not affiliated with or endorsed by Anthropic or OpenAI.
Built for people who look at the bill
Every part of the product assumes you care what this costs — because you should.
All-Inclusive Token Prices
Hosted open-source models at one transparent per-token price. No platform fee, no hidden markups — the price on the page is the price on the invoice.
0% BYOK Fee
Bring your own API keys and pay nothing — forever. Typical aggregators charge up to 5%. Your keys, your provider rates, our routing and analytics for free.
One Bill, Every Model
Unified billing across every provider and model — one balance, one statement, per-key spend visibility. Top up by card in under a minute.
Smart Routing & Failover
One OpenAI-compatible API in front of every provider. Requests route to the best available upstream and fail over automatically — no code changes, no downtime.
Stop overpaying for tokens
Sign up in a minute, top up only what you need, and pay for exactly what you use. No subscription to cancel — because there is no subscription.
Create Free Account