Supported Models
Tokensmart supports nearly every major LLM on the market. See the models page or pricing page for the complete list and real-time rates.
Major providers
OpenAI
gpt-5.5(flagship)gpt-5.4,gpt-5.4-minigpt-image-2(image generation, per-call billing)
Anthropic
claude-fable-5(Mythos-class, newest and strongest)claude-opus-4-8(Opus flagship)claude-opus-4-7,claude-opus-4-6claude-sonnet-4-6(best value)claude-haiku-4-5-20251001(fastest)
gemini-3-pro-preview(flagship)gemini-3.5-flashgemini-2.5-pro,gemini-2.5-pro-thinking
DeepSeek
deepseek-v4-pro(flagship: code / math / complex reasoning)deepseek-v4-flash(economy: faster, cheaper, high-concurrency)
Chinese models
qwen3.5-plus,qwen3.5-flashqwen3-max,qwen3-max-previewqwen3-vl-plus(vision)
Image generation (per-image billing)
gpt-image-2— $0.02 per image
⚠️ Image models use dedicated endpoints — do NOT send them to
/v1/chat/completions. See Image Generation API, Image Edits API, and Image Compose API.
Switching models
Just change the model field — no SDK swap, no new API key:
# Today: GPT
resp = client.chat.completions.create(model="gpt-5.5", messages=[...])
# Tomorrow: try Claude
resp = client.chat.completions.create(model="claude-sonnet-4-6", messages=[...])
New model launches
We track every provider's releases. New models typically land on Tokensmart within 24-48 hours of launch. Watch the changelog for updates.