Skip to main content

Model profile

Mellum2-12B-A2.5B-Thinking

JetBrainsCurrentReleased May 28, 2026
Data verified
Overall Score
Unranked
Arena Elo
Not listed
Eligible category ranks
2of 8
Price (1M tokens)
Not listedAPI pricing
Speed
Not listed
Context
128K

Evidence coverage

5 of 323 tracked benchmarks are published. 5 are verified and 0 provisional. 3 of 8 categories are measured.

Updated July 23, 2026Methodology
Published / tracked
5 / 323
Verified
5
Provisional
0
Categories with evidence
3 / 8

Evidence by category

  • Agentic1 benchmark
    Verified
  • Coding0 benchmarks
    Not measured
  • Reasoning0 benchmarks
    Not measured
  • Knowledge3 benchmarks
    Verified
  • Math0 benchmarks
    Not measured
  • Multilingual0 benchmarks
    Not measured
  • Multimodal0 benchmarks
    Not measured
  • Inst. Following1 benchmark
    Verified
Open WeightSelf-hostReasoning
Confidence:
Low
thinking

BenchLM is tracking Mellum2-12B-A2.5B-Thinking, but this profile is currently excluded from the public leaderboard because it still lacks enough non-generated benchmark coverage to rank safely. Only non-generated public benchmark rows appear below.

Mellum2-12B-A2.5B-Thinking is a open weight model with a 128K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

Mellum2-12B-A2.5B-Thinking sits inside the Mellum2 12B-A2.5B family alongside Mellum2-12B-A2.5B-Instruct. This profile currently has 5 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Instruction Following (#27), while its weakest is Agentic (#62). This performance profile makes it a well-rounded choice across a range of tasks.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 77.4483.93

  1. Claude Mythos 5
    Anthropic
    #183.93
    Claude Mythos 5 is #1 with a score of 83.93.
    Compare
  2. Claude Fable 5
    Anthropic
    #283.68
    Claude Fable 5 is #2 with a score of 83.68.
    Compare
  3. GPT-5.6 Sol
    OpenAI
    #381.96
    GPT-5.6 Sol is #3 with a score of 81.96.
    Compare
  4. Kimi K3
    Moonshot AI
    #480.96
    Kimi K3 is #4 with a score of 80.96.
    Compare
  5. Claude Opus 4.8
    Anthropic
    #578.34
    Claude Opus 4.8 is #5 with a score of 78.34.
    Compare
  6. Muse Spark 1.1
    Meta
    #677.44
    Muse Spark 1.1 is #6 with a score of 77.44.
    Compare
  7. Mellum2-12B-A2.5B-ThinkingCurrent model
    JetBrains
    UnrankedNot measured
    Mellum2-12B-A2.5B-Thinking is Unranked with a score of Not measured.

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Inst. Following13%
    Eligible cohort rank #27 of 31Category score 39.5
  2. Agentic48%
    Eligible cohort rank #62 of 119Category score 47.5

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #62 of 119Percentile 48thWeight 22%1 benchmarkVerified47.5
CodingWeight 20%0 benchmarksNot measuredNot measured
ReasoningWeight 17%0 benchmarksNot measuredNot measured
KnowledgeRank Not rankedWeight 12%3 benchmarksVerified46.4
MathWeight 5%0 benchmarksNot measuredNot measured
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalWeight 12%0 benchmarksNot measuredNot measured
Inst. FollowingRank #27 of 31Percentile 13thWeight 5%1 benchmarkVerified39.5

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic1 benchmark
BFCL v4Provider exact

Berkeley Function Calling Leaderboard v4

45.6%Display only
Source: JetBrains: Mellum2 Thinking model cardProvenance: Provider exact
Knowledge3 benchmarks
GPQAProvider exact

Graduate-Level Google-Proof Q&A

57.6%Weighted 7%
Source: JetBrains: Mellum2 Thinking model cardProvenance: JetBrains reports GPQA Diamond at 57.6. BenchLM stores that exact value on the weighted GPQA lane and the display GPQA-Diamond lane.
MMLU-ReduxProvider exact
86.2%Display only
Source: JetBrains: Mellum2 Thinking model cardProvenance: Provider exact
GPQA-DProvider exact

GPQA Diamond

57.6%Display only
Source: JetBrains: Mellum2 Thinking model cardProvenance: JetBrains reports GPQA Diamond at 57.6.
Inst. Following1 benchmark
IFEvalProvider exact

Instruction-Following Eval

76.5%Weighted 35%
Source: JetBrains: Mellum2 Thinking model cardProvenance: Provider exact

Mellum2 12B-A2.5B Family

Thinking

Frequently Asked Questions

How does Mellum2-12B-A2.5B-Thinking perform overall in AI benchmarks?

Mellum2-12B-A2.5B-Thinking has 5 published benchmark scores on BenchLM, but it does not yet have enough non-generated coverage to receive a global overall rank.

Is Mellum2-12B-A2.5B-Thinking good for knowledge and understanding?

Mellum2-12B-A2.5B-Thinking has visible benchmark coverage in knowledge and understanding, but BenchLM does not currently assign it a global category rank there.

Is Mellum2-12B-A2.5B-Thinking good for agentic tool use and computer tasks?

Mellum2-12B-A2.5B-Thinking ranks #62 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 47.5. There are stronger options in this category.

Is Mellum2-12B-A2.5B-Thinking good for instruction following?

Mellum2-12B-A2.5B-Thinking ranks #27 out of 31 models in instruction following benchmarks with an average score of 39.5. There are stronger options in this category.

Is Mellum2-12B-A2.5B-Thinking open source?

Yes, Mellum2-12B-A2.5B-Thinking is an open weight model created by JetBrains, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Which sibling models are related to Mellum2-12B-A2.5B-Thinking?

Mellum2-12B-A2.5B-Thinking belongs to the Mellum2 12B-A2.5B family. Related variants on BenchLM include Mellum2-12B-A2.5B-Instruct.

Does Mellum2-12B-A2.5B-Thinking have full benchmark coverage on BenchLM?

Not yet. Mellum2-12B-A2.5B-Thinking currently has 5 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of Mellum2-12B-A2.5B-Thinking?

Mellum2-12B-A2.5B-Thinking has a published context window of 128K, which determines how much text it can process in a single interaction.

Last updated: July 23, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.