Home/Models/MiMo-V2.5
1M contextReasoningResponses API

MiMo-V2.5 on QuickSilver Pro

MiMo-V2.5 is Xiaomi's open-weight coding and agent model with a 1M-token context. QuickSilver Pro is $0.112 input / $0.224 output per 1M tokens — 20% below Xiaomi's $0.14 / $0.28 official route and equal to OpenRouter's lowest current endpoint. It thinks by default; pass `reasoning: { enabled: false }` for direct output.

$0.112 input · $0.224 output per 1M tokens
ByRaullen Chai·Updated

At a glance

Context
1M tokens
Input / 1M
$0.112
Output / 1M
$0.224
Thinks by default
Yes

Cost-efficient open-weight coding and agents over a 1M-token context.

Pricing comparison ($/1M tokens)

ProviderInputOutputvs QSP
QuickSilver Pro$0.112$0.224lowest-cost
OpenRouter (xiaomi/mimo-v2.5)$0.14$0.2820% lower

When to use

Reach for MiMo-V2.5 for cost-efficient coding and agentic workflows, batch codegen, and long-context tasks that need a 1M-token working set. Xiaomi ships it open-weight, so it suits everyday coding loops that do not need a frontier reasoner.

When to use something else

For frontier reasoning, complex multi-file refactors, or the hardest agentic tasks, step up to MiniMax M3, DeepSeek V4 Pro, or Kimi K3. For pure mathematical reasoning, DeepSeek V4 Pro. If you don't need the 1M context, DeepSeek V4 Flash ($0.112/$0.224) is close in price on shorter contexts.

Quickstart (curl)

curl https://api.quicksilverpro.io/v1/chat/completions \
  -H "Authorization: Bearer $QSP_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mimo-v2.5",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

OpenAI-compatible. One-line migration via base_url.

FAQ

MiMo-V2.5 is a low-cost open-weight coding and agent model at $0.112 input / $0.224 output per 1M tokens. Qwen 3.7 Flash is now lower for fast multimodal work; MiMo remains useful when its coding behavior performs better on your evals.

Yes — MiMo-V2.5 is an OpenAI-compatible chat completions endpoint on QuickSilver Pro. Set base_url=https://api.quicksilverpro.io/v1, paste your QSP key, and use model="mimo-v2.5". Streaming, tool calling, json_schema strict mode, and usage.cost accounting all work. It thinks by default; pass `reasoning: { enabled: false }` for non-thinking output.

QuickSilver Pro is $0.112 / $0.224 per 1M tokens — 20% below Xiaomi's official $0.14 / $0.28 OpenRouter route and equal to OpenRouter's lowest current provider endpoint. We therefore do not claim to beat every possible OpenRouter route. Drop the `xiaomi/` prefix when migrating.

Try MiMo-V2.5 with double credits — up to $50 in bonus credits

Get API Key