MiMo-V2.6-Pro
Xiaomi · released
Xiaomi's trillion-parameter open-weight flagship, taking text, images, audio and video as input. It is the strongest open-weights model on the Intelligence Index and among the cheapest capable models to run.
Best for: Open-weights frontier · Low-cost agentic work · Multimodal input
- Intelligence Index
- 46.3
- #9 of 38
- Price per 1M tokens
- $0.43 in $0.87 out
- Blended $0.54 at 3:1
- Context window
- 1M
- Up to 131K output
- Output speed
- 38 tokens/s
- $0.13 measured cost per task
Capability profile
One 0–100 score per area, placing this model between the weakest and strongest tracked model across that area's benchmarks.
- Overall74 · #8 of 38
Based on 3 of 3 benchmarks
- Coding67 · #10 of 39
Based on 5 of 5 benchmarks
- Agents & knowledge work73 · #6 of 40
Based on 6 of 6 benchmarks
- Reasoning & science80 · #7 of 39
Based on 2 of 2 benchmarks
- Knowledge & reliability73 · #13 of 39
Based on 2 of 2 benchmarks
- Vision45 · #26 of 31
Based on 1 of 2 benchmarks
Benchmark results
All third-party results on record. Rank is among the tracked models that have been tested on that benchmark.
Overall
#9 of 38· Artificial Analysis
- Vals Index55.2%
#8 of 26· $0.41 per test· Vals AI
- Arena · Text1480 ±9
#10 of 33· 4,056 votes· Arena
Coding
- Terminal-Bench 4.034.8%
#10 of 38· Artificial Analysis
- SciCode60.9%
#5 of 38· Artificial Analysis
- Arena · WebDev1630 ±11
#14 of 34· 3,415 votes· Arena
- Vibe Code Bench85.2%
#10 of 26· Vals AI
- Code Migration42.3%
#15 of 26· Vals AI
Agents & knowledge work
- GDPval-AA59.3
#5 of 39· Artificial Analysis
- Arena · Agent+3.3 ±1.7
#12 of 29· Arena
- Finance Agent v257.3%
#7 of 26· Vals AI
#14 of 26· Vals AI
- Legal Research Bench47.1%
#8 of 26· Vals AI
- Tax Agent Bench64.9%
#12 of 26· Vals AI
Reasoning & science
- Humanity's Last Exam49.4%
#7 of 39· Artificial Analysis
- CritPt26.6%
#8 of 39· Artificial Analysis
- ARC-AGI-2saturatingNot tested
- ARC-AGI-3Not tested
- GPQA DiamondsaturatingNot tested
Knowledge & reliability
#16 of 39· Artificial Analysis
- AA-LCR86.3%
#3 of 39· Artificial Analysis
Vision
- MMMU-ProNot tested
- Arena · Vision1245 ±15
#15 of 19· 1,787 votes· Arena
Specifications
- Lab
- Xiaomi
- Released
- Sep 22, 2026
- Availability
- Generally available
- Context window
- 1,048,576 tokens
- Max output
- 131,072 tokens
- Accepts
- Text, Images, Audio, Video
- Knowledge cutoff
- Not published
- Weights
- Open
- Licence
- MIT
- Parameters
- 1.02T total · 42B active
- API model ID
mimo-v2.6-pro
Pricing
- Input / 1M tokens
- $0.43
- Cached input / 1M
- <$0.01
- Output / 1M tokens
- $0.87
- Blended (3:1)
- $0.54
- Measured cost / task
- $0.13
Worth knowing
- Release details come from third-party coverage and the model listings; check the model card before relying on the licence terms.
Official sources
Closest alternatives
The models nearest to this one on overall score.
23 benchmarks tracked. See methodology and sources.