GPQA DiamondVerified
87.9%
GPQA Diamond (AA run)
Models / Moonshot
Kimi K2.5
Kimi · Released 2026-01-27
Moonshot's January 2026 multimodal open-weights model. 262K context, $0.60/$3 per 1M tokens. Agent Swarm up to 100 sub-agents. Predecessor to K2.6.
Open-weights Moonshot pick when K2.6 isn't available or you need the K2.5 weight checkpoint.
Composite60
Context262K
Input / 1M$0.60
Output / 1M$3.0
Knowledge cutoff2025-10-01
Output speed57 t/s
Statussolid
Family profile
Best published score in each covered benchmark family.
71/100 avg
- Reasoning
- 88
- Agentic
- 46
- Long context
- 78
3 tested benchmark families
Benchmark placements
Per-benchmark positions
Where Kimi K2.5 places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.
| Benchmark | Family | Rank | Score | Source | Date |
|---|---|---|---|---|---|
| AA-LCR | Long context | #65/ 184 | 78 | Artificial Analysis | 2026-09-28 |
| GPQA Diamond | Reasoning | #66/ 184 | 88 | Artificial Analysis | 2026-09-28 |
| Humanity's Last Exam | Reasoning | #79/ 186 | 31 | Artificial Analysis | 2026-09-28 |
| Terminal-Bench | Agentic | #82/ 167 | 46 | Artificial Analysis | 2026-09-28 |
| AA output speed | Performance | #107/ 195 | Artificial Analysis | 2026-10-12 |
Family context
Strongest rows in each family
The three highest-scoring models with pages in each capability family. Where Kimi K2.5 shows up, it's highlighted.
ReasoningKimi K2.5 · 88
- 196
- 295
- 395
Math
- 152
- 252
- 350
AgenticKimi K2.5 · 46
- 191
- 289
- 389
Long contextKimi K2.5 · 78
- 189
- 286
- 385
Newest receipts
5 recentLast 120 days
Benchmark rows added to the public ledger for Kimi K2.5 in the last 120 days. Older rows live in the full table below.
| Benchmark | Family | Score | Source | Days ago |
|---|---|---|---|---|
| AA output speed | Performance | 57 t/s | Artificial Analysis | 10 |
| AA-LCR | Long context | 78% | Artificial Analysis | 24 |
| GPQA Diamond | Reasoning | 87.9% | Artificial Analysis | 24 |
| Humanity's Last Exam | Reasoning | 30.7% | Artificial Analysis | 24 |
| Terminal-Bench | Agentic | 45.69% | Artificial Analysis | 24 |
Family coverage
Best score per family
A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.
Price vs. performance
Where Kimi K2.5 sits on the scatter
Hover any dot for name, score, and input price. Keyboard: tab through the top twelve, or use the ranking below.
AnthropicOpenAIGoogleAlibabaMetaZhipuxAIDeepSeekMiniMaxXiaomiMoonshotOtherMeituanMistralNVIDIACohere
Full benchmark ledger
Every catalog benchmark for Kimi K2.5. Scores link to the original source; gaps mean no public row exists yet.
5 of 13 catalog benchmarks have a sourced row for Kimi K2.5.
Reasoning
Humanity's Last ExamVerified
30.7%
HLE (AA run)
Agentic
Terminal-BenchVerified
45.69%
Terminal-Bench v2.1 (AA run)
Long context
AA-LCRVerified
78%
AA-LCR (AA run)
Performance
AA output speedVerified
57 t/s
Median output tokens/s (1k prompt, default provider)
Sources
Where the numbers come from
- Moonshot API pricing· checked 2026-06-09
- Kimi K2.5 on Hugging Face· checked 2026-06-09
Questions this model page answers
Every answer below is assembled from the dated fields on this page. Nothing is written separately for search.
What is Kimi K2.5?
Kimi K2.5 is a Moonshot model: Moonshot's January 2026 multimodal open-weights model. 262K context, $0.60/$3 per 1M tokens. Agent Swarm up to 100 sub-agents. Predecessor to K2.6.
What is Kimi K2.5 for?
Open-weights Moonshot pick when K2.6 isn't available or you need the K2.5 weight checkpoint.
What does Kimi K2.5 cost?
Kimi K2.5 lists $0.60 input and $3.0 output per 1M tokens. Pricing last checked 2026-06-09.
How large is the Kimi K2.5 context window?
Kimi K2.5 publishes a 262K token context window.
What is the VerdictPal status of Kimi K2.5?
Kimi K2.5 carries Solid status on the model atlas.
The Pack · Editorial newsletter
Know when the rules move.
One short email when a university changes its AI rule for theses, or a research tool changes its pricing or privacy terms. No tracking, no third-party analytics. Unsubscribe in one click.