Models available on BetterToken: updates and Model IDs

Check the model plaza for currently available models, their exact Model IDs, and current prices. Use the announcements below to compare suitable tasks and find setup guidance; confirm availability in the model plaza before configuring your tool.

Claude models

Model update	Model ID	Suitable tasks
Claude Haiku 5.5	`claude-haiku-5-5`	Classification, extraction, summaries, customer support, and focused subagent tasks.
Claude Sonnet 5.5	`claude-sonnet-5-5`	Well-defined coding tasks, bug fixes, documents, slides, and spreadsheets.
Claude Opus 5.5	`claude-opus-5-5`	Complex repository changes, audits, and long-running coding agents.
Claude Fable 5.1	`claude-fable-5-1`	Demanding reasoning, long-running coding agents, multistep research, and complex document work.
Claude Opus 5	`claude-opus-5`	Complex coding and multi-step work across large codebases.
Claude Sonnet 5	`claude-sonnet-5`	Long-context coding, automatic fixes, code review, and multi-step agent workflows.
Claude Fable 5	See model plaza	Long-running agents, complex software engineering, and intensive knowledge work.
Claude Opus 4.7	See model plaza	Advanced software engineering, complex long-running tasks, and professional knowledge work.

GPT and Codex models

Model update	Model ID	Suitable tasks
GPT-6.1 Sol	`gpt-6.1-sol`	Multi-file changes, debugging, and longer software-engineering tasks.
GPT-6 Sol and GPT-6 Luna	Sol: `gpt-6-sol`; Luna: `gpt-6-luna`	Use Sol for complex coding and agentic workflows, or Luna for focused, repeatable, high-volume tasks.
GPT-6 Astra	`gpt-6-astra`	Difficult coding, research, computer use, and long-running agent tasks.
GPT-5.6 family	Sol: `gpt-5.6-sol`; Terra: `gpt-5.6-terra`; Luna: `gpt-5.6-luna`	Use Sol for complex reasoning and large codebases, Terra for everyday coding, or Luna for lightweight Q&A and batch assistance.
GPT-5.5	See model plaza	Multi-step workflows combining implementation, debugging, validation, and documentation.

Image models

Model update	Model ID	Suitable tasks
GPT Image 2.5	Flare: `gpt-image-2.5-flare`; Sunburst: `gpt-image-2.5-sunburst`	Use Flare for visual drafts and everyday editing, or Sunburst for product images, brand assets, and precise final artwork.
GPT Image 2	`gpt-image-2`	Product launch posters, campaign assets, social images, and visual concept drafts.
Nano Banana family	Pro: `nano-banana-pro`; 2: `nano-banana-2`; 2.1: `nano-banana-2.1`	Use Pro for complex compositions and precise final artwork, 2 for quick drafts and everyday editing, or 2.1 for posters with text, reference-image blending, and multi-turn edits.

DeepSeek models

Model update	Model ID	Suitable tasks
DeepSeek V4.1 Flash 1017	`deepseek-v4.1-flash-1017`	Repository changes, debugging, and building projects from a specification.
DeepSeek V4.1 Flash	`deepseek-flash`	Evaluate coding changes, agent workflows, debugging, planning, and analysis.

MiniMax models

Model update	Model ID	Suitable tasks
MiniMax M3.1 Flash	`MiniMax-M3.1-Flash`	Implementing features, debugging, and processing documents across several steps.

How to switch models in your tool

Use your BetterToken API Key and copy the exact Model ID from the model plaza into your client’s model setting. Choose the Base URL for your protocol: https://bettertoken.ai for Claude Code or Anthropic Messages, and https://bettertoken.ai/v1 for Codex CLI or OpenAI-compatible clients.

Model selection guide: compare models by task, quality, speed, and context length.
Verify model identity: check which model was used through request records and response fields.
Switch models in Claude Code: update your configuration and restart, or use /model YOUR_MODEL_ID during a session.
Supported tools and setup guides: find your tool’s authentication, Base URL, and model configuration steps.

Common questions

When can I use a new model in my tool?

Confirm that the model is available in the model plaza, then configure its Model ID using your tool’s setup guide. If your client does not refresh its model list automatically, restart it and send a short test request. Check the matching request in the Dashboard.

Does Model ID capitalization matter?

Copy the Model ID exactly, preserving capitalization and punctuation. For example, use MiniMax-M3.1-Flash for MiniMax M3.1 Flash. A display name or an ID from another platform may differ from the BetterToken Model ID.

Can Claude Desktop use every model listed here?

Claude Desktop supports only Claude models because of a client limitation. One BetterToken API Key works with all available models; Claude Code can call GPT and other available models through Anthropic Messages. See the Claude Desktop setup guide.

​Claude models

​GPT and Codex models

​Image models

​DeepSeek models

​MiniMax models

​How to switch models in your tool

​Common questions

​When can I use a new model in my tool?

​Does Model ID capitalization matter?

​Can Claude Desktop use every model listed here?