Pricing
One simple, credit-based price.
Every model - across every provider - is billed at a flat 1 credit per 1,000 tokens. No per-model math, no surprises. New accounts start with 50 free credits.
| Model | Context | Price | Max Output | Latency |
|---|---|---|---|---|
 OpenAI | ||||
GPT-4oopenai/gpt-4o | 128K | 1 cr / 1K tok | 16,384 | ~0.5s |
GPT-4o Miniopenai/gpt-4o-mini | 128K | 1 cr / 1K tok | 16,384 | ~0.3s |
.svg){kind=icon} Anthropic | ||||
Claude 3.5 Sonnetanthropic/claude-3-5-sonnet | 200K | 1 cr / 1K tok | 8,096 | ~0.8s |
Claude 3 Haikuanthropic/claude-3-haiku | 200K | 1 cr / 1K tok | 4,096 | ~0.2s |
 Google | ||||
Gemini 2.0 Flashgoogle/gemini-2.0-flash | 1M | 1 cr / 1K tok | 8,192 | ~0.3s |
Gemini 1.5 Progoogle/gemini-1.5-pro | 2M | 1 cr / 1K tok | 8,192 | ~1.2s |
 Meta | ||||
Llama 3.3 70Bmeta/llama-3.3-70b-instruct | 128K | 1 cr / 1K tok | 8,192 | ~0.6s |
 Mistral AI | ||||
Mistral Largemistralai/mistral-large | 128K | 1 cr / 1K tok | 8,192 | ~0.9s |
 DeepSeek | ||||
DeepSeek R1deepseek/deepseek-r1 | 64K | 1 cr / 1K tok | 8,192 | ~2.1s |
 Alibaba | ||||
Qwen 2.5 72Balibaba/qwen-2.5-72b-instruct | 128K | 1 cr / 1K tok | 8,192 | ~0.5s |
Qwen Maxqwen-max | 32K | 1 cr / 1K tok | 8,192 | ~0.8s |
Qwen Plusqwen-plus | 131K | 1 cr / 1K tok | 8,192 | ~0.5s |
Qwen3 235Bqwen3-235b-a22b | 131K | 1 cr / 1K tok | 8,192 | ~0.9s |
 xAI | ||||
Grok 3 Betaxai/grok-3-beta | 131K | 1 cr / 1K tok | 8,192 | ~1.0s |
 Cohere | ||||
Command R+cohere/command-r-plus | 128K | 1 cr / 1K tok | 4,096 | ~1.1s |
Cerebras | ||||
Llama 3.3 70Bllama-3.3-70b | 128K | 1 cr / 1K tok | 8,192 | ~0.1s |
Qwen 3 32Bqwen-3-32b | 128K | 1 cr / 1K tok | 8,192 | ~0.1s |
GPT-OSS 120Bgpt-oss-120b | 128K | 1 cr / 1K tok | 8,192 | ~0.1s |
Llama 4 Scoutllama-4-scout-17b-16e-instruct | 128K | 1 cr / 1K tok | 8,192 | ~0.1s |
NVIDIA | ||||
Llama 4 Maverickmeta/llama-4-maverick-17b-128e-instruct | 1M | 1 cr / 1K tok | 8,192 | ~0.7s |
DeepSeek R1deepseek-ai/deepseek-r1 | 128K | 1 cr / 1K tok | 8,192 | ~2.0s |
Qwen3 Coder 480Bqwen/qwen3-coder-480b-a35b-instruct | 256K | 1 cr / 1K tok | 8,192 | ~1.2s |
Llama 3.3 Nemotron Super 49Bnvidia/llama-3.3-nemotron-super-49b-v1 | 128K | 1 cr / 1K tok | 8,192 | ~0.8s |
OpenAIAnthropicGoogleMetaMistral AIDeepSeekAlibabaxAICohereCerebras
NVIDIA
Pricing is credit-based: 1 credit = 1,000 tokens (input + output combined), billed at a flat rate across every model and provider. New accounts get 50 complimentary credits.