GPQA
87.6 score
LLM Leaderboard · Proprietary
Kimi K2.5 (Reasoning) benchmarks, pricing, and LLM comparison.
Compare Kimi K2.5 (Reasoning) vs GPT, Claude, Gemini, DeepSeek, open-weight, and frontier AI models using public benchmark scores, token pricing, context window, and access details.
Rank #33AskClash overall score: 18.0
$0.60 / $3.00Input and output token price, when published. Context: 256K.
APIBilling and access path cached for this model row.
Kimi K2.5 (Reasoning) benchmark snapshot
AskClash combines public LLM benchmark cells into a weighted percentile score and penalizes missing coverage so narrow rows do not dominate better-measured models.
Overall18.0
Benchmark cells4
Context256K
CreatorMoonshot AI
Kimi K2.5 (Reasoning) public benchmark scores
Cached benchmark values can include HLE, GPQA, SWE-bench, SWE-Pro, SWE-Atlas, Terminal-Bench, MCP Atlas, MMMU-Pro, ARC-AGI-2, Tau2, and model-specific coding or agent scores.
SWE-bench
76.8 score
MMMU-Pro
78.5 score
Tau2
95.9 score
Kimi K2.5 (Reasoning) vs other AI models
Use these comparison links to evaluate Kimi K2.5 (Reasoning) against nearby LLMs by benchmark score, price, context window, and provider.
Kimi K2.5 (Reasoning) vs Claude Mythos/Fable 5Kimi K2.5 (Reasoning) vs Claude Opus 4.8 (Adaptive)Kimi K2.5 (Reasoning) vs GPT-5.5 xHighKimi K2.5 (Reasoning) vs Claude Opus 4.7 (Adaptive)Kimi K2.5 (Reasoning) vs GPT-5.4 xHighKimi K2.5 (Reasoning) vs Gemini 3.5 Flash HighKimi K2.5 (Reasoning) vs Kimi K2.7 CodeKimi K2.5 (Reasoning) vs Gemini 3.1 Pro
Related AI and tech coverage
Cached AskClash article matches that can provide release, provider, benchmark, pricing, or market context around this model.
Moonshot AI releases Kimi K2.7-Code, claiming 30% lower reasoning token usage compared to K2.6, available under a modified MIT license (Sean Michael Kerner/VentureBeat)
Techmeme
HuggingFace Transformers v5.5.4 Release Notes
** Fix Kimi-K2.5 tokenizer regression and _patch_mistral_regex Attribute… (#45305) by ArthurZucker ** Fix Qwen2.5-VL temporal RoPE scaling applied to still images (#45330) by Kash6, zucchini-nlp
[AINews] Moonshot Kimi K2.6: the world's leading Open Model refreshes to catch up to Opus 4.6 (ahead of DeepSeek v4?)
Latent Space
AgentFlow specs.py — Core Data Models (Pydantic Specs for Pipelines, Nodes, Agents)
Pydantic data models defining AgentFlow's type system: AgentKind (codex/claude/kimi/python/shell/sync), NodeSpec, PipelineSpec, ProviderConfig, target types (local/SSH/EC2/ECS/container), fanout expansion, MCP server specs.
Last cached leaderboard date: June 9, 2026. This model page is generated from the AskClash LLM Leaderboard cache and linked from the live leaderboard.