K2-V2 (high)
K2-V2 (high) is behind the leaders in instruction following. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, or factuality.
Capability profile
Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.
Too few results yet to rate Reasoning, Coding, Agentic, Safety, Long Context, Math, Multimodal, Multilingual or Factuality.
Price
No current price is tracked for this model. See the rate card
Evidence
34results on12benchmarks
- 34 aggregator
From 1 source · latest Oct 8, 2026 · How verification works
K2-V2 (high) benchmark results
34 results on 12 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.
33.8% behind the leader1 of 3 ranked benchmarks measured
- IFBench60.14Oct 8, 2026aa_ifbench
0 of 6 ranked benchmarks measured
- 0.00Oct 8, 2026
- GPQA Diamond59.80Oct 8, 2026gpqa
- GPQA Diamond54.14Oct 8, 2026gpqa
Show 4 more reasoning resultsHide 4 reasoning results
- GPQA Diamond68.08Oct 8, 2026gpqa
- Humanity's Last Exam3.61Oct 8, 2026aa_hle
- Humanity's Last Exam4.36Oct 8, 2026aa_hle
- Humanity's Last Exam10.52Oct 8, 2026aa_hle
0 of 10 ranked benchmarks measured
- Terminal-Bench Hard4.55Oct 8, 2026aa_terminalbench_hard
- Terminal-Bench Hard8.33Oct 8, 2026aa_terminalbench_hard
- Terminal-Bench Hard9.85Oct 8, 2026aa_terminalbench_hard
0 of 3 ranked benchmarks measured
- 20.33Oct 8, 2026
- 28.00Oct 8, 2026
- 35.00Oct 8, 2026
0 of 4 ranked benchmarks measured
- AA-Omniscience · Accuracy16.33Oct 8, 2026omniscienceAccuracy
- AA-Omniscience · Accuracy17.93Oct 8, 2026omniscienceAccuracy
- AA-Omniscience · Non-hallucination23.96Oct 8, 2026omniscienceNonHallucination
Show 3 more factuality resultsHide 3 factuality results
- AA-Omniscience · Accuracy18.53Oct 8, 2026omniscienceAccuracy
- AA-Omniscience · Non-hallucination18.44Oct 8, 2026omniscienceNonHallucination
- AA-Omniscience · Non-hallucination9.02Oct 8, 2026omniscienceNonHallucination
More results
Benchmarks outside the capability baskets. They are not ranked against a leader.
- AA Intelligence7.31Oct 8, 2026aa_intelligence_index
- -47.28Oct 8, 2026
- τ²-Bench Telecom (AA run)20.76Oct 8, 2026aa_tau2
- AA Intelligence9.02Oct 8, 2026aa_intelligence_index
- -49.00Oct 8, 2026
- τ²-Bench Telecom (AA run)24.85Oct 8, 2026aa_tau2
Show 3 more resultsHide 3 results
- AA Intelligence9.87Oct 8, 2026aa_intelligence_index
- -55.58Oct 8, 2026
- τ²-Bench Telecom (AA run)27.78Oct 8, 2026aa_tau2
K2-V2 (high): common questions
Who makes K2-V2 (high)?
K2-V2 (high) is made by MBZUAI.
When was K2-V2 (high) released?
K2-V2 (high) was released on Dec 5, 2025, according to Artificial Analysis.
What is K2-V2 (high) good at?
K2-V2 (high) is behind the leaders in instruction following. Too few results yet to rate reasoning, coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, or factuality.
How many benchmarks has K2-V2 (high) been tested on?
We track 34 results for K2-V2 (high) on 12 benchmarks from 1 source. The latest was recorded on Oct 8, 2026.
About this record
Where K2-V2 (high)'s numbers come from, and every name it appears under.
- Tracked since
- Jun 29, 2026
- Newest source mention
- Jun 29, 2026
Where the results come from
Verification: 34 scores · 0 independently verified · 34 aggregator-attributed. How these tiers are assigned
From 1 source on 1 site. Artificial Analysis supplies all of them. Bars are coloured by trust tier.
- artificialanalysis.ai34
Also known as
How our sources name K2-V2 (high) at each reasoning setting.
| Setting | API id |
|---|---|
| low | k2-v2 (low) |
| medium | k2-v2 (medium) |