feat: add MiniMax model support with M3 as default - #354
Conversation
- Add MiniMax-M2.7 and MiniMax-M2.7-highspeed to model list - Include MiniMax-M2.5 and MiniMax-M2.5-highspeed as alternatives - Add MiniMax provider routing with api.minimax.io/v1 base URL - Add MiniMax BYOK API key template - Register MiniMax in platform-enabled providers list
|
i wish they'd accept it |
- Add MiniMax-M3 to model list and set as default - Keep MiniMax-M2.7 and MiniMax-M2.7-highspeed - Remove older models (M2.5 and M2.5-highspeed) - Update README, .dev.vars.example, docs/llm.md, and worker-configuration.d.ts
|
Hi @octo-patch — a few things need to happen before CI can run on this PR:
Once those are in place CI should pick up automatically; let me know if it doesn't. |
Summary
Upgrade MiniMax model configuration in the M3 update.
Changes
MINIMAX_API_KEYin worker configuration typesMINIMAX_API_KEYin README, .dev.vars.example, and docs/llm.mdWhy
MiniMax-M3 is the latest model, with a 512K context window, 128K max output, and image input support. Making M3 the default gives users the strongest quality while keeping M2.7 available for cost-sensitive workloads.
Testing
worker/agents/inferutils/core.ts(case 'minimax'→https://api.minimax.io/v1)worker/types/secretsTemplates.ts