Xiaomi: MiMo-V2.5
by XiaomiXiaomi: MiMo-V2.5 is a large language model from Xiaomi. It costs $0.400 per million input tokens and $2.00 per million output tokens. Its context window is 1.1M tokens.
- Input / 1M tokens
- $0.400
- Output / 1M tokens
- $2.00
- Cached input / 1M
- $0.080
- Context window
- 1.1M
On repeated prefixes
tokens
Who serves it cheapest
2 hosts serve MiMo-V2.5. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| GMICloudCheapestfp8 | $0.119 | $0.238 | 1.1M | - | 0.0% |
| Venicefp8 | $0.400 | $2.00 | 1M | - | 97.1% |
The spread between GMICloud and Venice is 5.4× for identical weights. Quantization and context limits differ, so check both columns before switching.
Benchmarks
Independent scores published alongside the catalogue.
- Coding index
- 56.8
- Agentic index
- 15.8
About MiMo-V2.5
MiMo-V2.5 is a native omnimodal model by Xiaomi. It delivers Pro-level agentic performance at roughly half the inference cost, while surpassing MiMo-V2-Omni in multimodal perception across image and video understanding...
Specifications
| Model ID | xiaomi/mimo-v2.5 |
|---|---|
| Provider | Xiaomi |
| Context window | 1.1M tokens |
| Max output | 66K tokens |
| Input modalities | text, audio, image, video |
| Output modalities | text |
| Knowledge cutoff | - |
| Open weights | Yes - XiaomiMiMo/MiMo-V2.5 |
| Released | April 22, 2026 |
Cheaper alternatives
Models that cost less than MiMo-V2.5 while keeping at least half its context window and every input modality it supports.
Frequently asked
How much does Xiaomi: MiMo-V2.5 cost?
$0.400 per million input tokens and $2.00 per million output tokens. Cached input reads cost $0.080 per million tokens.
What is the context window of Xiaomi: MiMo-V2.5?
1.1M tokens, with up to 66K tokens of output per request.
Which provider serves Xiaomi: MiMo-V2.5 cheapest?
GMICloud at $0.119 per million input tokens - 5.4× cheaper than Venice, the most expensive of the 2 hosts serving it.