Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 1.5B/7B/8B/14B/32B/70B | 1T |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | — | 256k |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.49AtlasCloud |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $2.50AtlasCloud |
| — | $223AtlasCloud · int4 | |
| Benchmarks | ||
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | 57.5% | 85% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 65.2% | 87.6% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | — | 52% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | — | 76.8% |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | — | 18% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 60.6% |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | — | 30.1% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 94.5% | — |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | 70% | — |
OSWorld-VerifiedAgentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better. | — | 63.3% |
| Overview | ||
| Company | DeepSeek | Moonshot AI |
| Release date | Jan 20 2025 | Jan 27 2026 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek R1 DistillvsClaude Fable 5.2Kimi K2.5vsClaude Fable 5.2DeepSeek R1 DistillvsGPT-6.2 AstraKimi K2.5vsGPT-6.2 AstraDeepSeek R1 DistillvsGemini 4 HydrogenKimi K2.5vsGemini 4 HydrogenDeepSeek R1 DistillvsMuse EmberKimi K2.5vsMuse EmberDeepSeek R1 DistillvsGrok 4.7Kimi K2.5vsGrok 4.7DeepSeek R1 DistillvsMistral Large 4Kimi K2.5vsMistral Large 4Frequently asked questions
Kimi K2.5 leads DeepSeek R1 Distill on 2 of the 2 benchmarks they both report (LiveCodeBench, GPQA Diamond). DeepSeek R1 Distill shipped 372 days before Kimi K2.5, so benchmark comparisons should account for the intervening progress.
DeepSeek R1 Distill has 1.5B/7B/8B/14B/32B/70B parameters, while Kimi K2.5 has 1T.
On LiveCodeBench, Kimi K2.5 leads at 85% vs DeepSeek R1 Distill at 57.5%. On GPQA Diamond, Kimi K2.5 leads at 87.6% vs DeepSeek R1 Distill at 65.2%.