Claude Haiku 4.5 vs Kimi K2.5
At a Glance
| Compare | Claude Haiku 4.5Anthropic |  Kimi K2.5Moonshot AI |
|---|---|---|
| Intelligence, Cost, and Efficiency | ||
| IntelligenceHigher is better · MM Intelligence v2.5 | #43 of 469.0 score · 2/3 sources · provisional · missing LiveBench · full-core range 6.0–39.3 | UnrankedNot in the 46-model eligible cohort |
| Pricing and Limits | ||
| Input priceFrom · USD / 1M tokens | $1.00Anthropic ↗ · Sep 3, 2026 | $0.45Deepinfra ↗ · Aug 29, 2026 |
| Output priceFrom · USD / 1M tokens | $5.00Anthropic ↗ · Sep 3, 2026 | $2.25Deepinfra ↗ · Aug 29, 2026 |
| Context windowMaximum documented tokens | 200K | 262K |
| Model facts checked | Aug 28, 2026View model evidence → | Aug 28, 2026View model evidence → |
Token prices are the lowest available sourced USD rates; input and output may use different providers. Cost ranking estimates output spend on LiveBench, not a full request bill. Ranking methodology →
Available Benchmarks
| Benchmark | Claude Haiku 4.5 | Kimi-K2.5 |
|---|---|---|
| LMArena Document Arenadocument-2026-09-13-d25aabda0010 · arena_rating · statistical tie | 1,419.9799% of row best · rating · claude-haiku-4-5-20251001; 95% CI [1414.09579972, 1425.83505080]; votes 34677; rank 37 | 1,430.10100% of row best · rating · kimi-k2.5-thinking; 95% CI [1423.37027866, 1436.81983378]; votes 21569; rank 34 |
| LMArena Text Arenatext-2026-09-13-d25aabda0010 · arena_rating · leader | 1,396.6497% of row best · rating · claude-haiku-4-5-20251001; 95% CI [1394.12644647, 1399.14715328]; votes 129278; rank 147 | 1,445.64100% of row best · rating · kimi-k2.5-thinking; 95% CI [1442.30343195, 1448.97544647]; votes 70513; rank 53 |
| Overall ResultCounted from the protocol-matched rows above · 1 tie | 0 benchmark winsNo overall winner | 1 benchmark winNo overall winner |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Side-by-Side Facts
| Field | Claude Haiku 4.5 | Kimi-K2.5 |
|---|---|---|
| Developer | Anthropic | Moonshot AI |
| Family | Claude 4 5 | Kimi K2 5 |
| Model | Claude Haiku 4.5 | Kimi-K2.5 |
| Version | Claude Haiku 4.5 | Kimi-K2.5 |
| Lifecycle | active | active |
| Released | 2025-10-15 | 2026-01-27 |
| Knowledge cutoff | Unknown | Unknown |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
| Context window | 200K | 262K |
| Total parameters | Unknown | 1T |
| Active parameters | Unknown | 32B |
| License | Unknown | other |
| Open weights | No | Yes |
| API available | Yes | Yes |
| Self-hostable | No | Yes |
| Provider access | Anthropic (Standard), Deepinfra (Standard) | Deepinfra (Standard), Hugging Face (Standard), Openrouter (Standard) |
| Capabilities | chat, generation, reasoning, tools | chat, generation, reasoning, tools |
Claude Haiku 4.5 Capabilities
Kimi K2.5 Capabilities
Related Comparisons
All image comparisons →| A | Pair | B | Context |
|---|---|---|---|
Claude Haiku 4.5Anthropic | vs |  GPT-6 AstraOpenAI | cross-developer peersimage, text |
Claude Fable 5.1Anthropic | vs | Claude Haiku 4.5Anthropic | family variantsimage, text |
Claude Haiku 4.5Anthropic | vs |  Gemini 3.1 ProGoogle DeepMind | cross-developer peerstext |
Claude Haiku 4.5Anthropic | vs |  DeepSeek-V4.1-FlashDeepSeek | cross-developer peersimage, text |
Claude Haiku 4.5Anthropic | vs | Grok 4.6xAI | cross-developer peersimage, text |
Claude Haiku 4.5Anthropic | vs |  Qwen3.8-MaxQwen | cross-developer peersimage, text |
Claude Haiku 4.5Anthropic | vs | Kimi-K3Moonshot AI | cross-developer peersimage, text |
 MiniMax-M3MiniMax | vs | Claude Haiku 4.5Anthropic | cross-developer peersimage, text |
Claude Haiku 4.5Anthropic | vs |  GLM-5.3Z.ai | cross-developer peerstext |
Claude Haiku 4.5Anthropic | vs |  Hy4 previewTencent | cross-developer peerstext |
Claude Haiku 4.5Anthropic | vs | Seed 2.1 ProByteDance Seed | cross-developer peersimage, text |
Claude Haiku 4.5Anthropic | vs |  Mistral Large 3Mistral AI | cross-developer peersimage, text |
Primary Evidence
Sources and Freshness
Questions
Claude Haiku 4.5 vs Kimi K2.5 FAQs
Is Claude Haiku 4.5 or Kimi K2.5 better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both Claude Haiku 4.5 and Kimi K2.5, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, Claude Haiku 4.5 or Kimi K2.5?+
Claude Haiku 4.5 is $1.00 and Kimi K2.5 is $0.45 per million tokens, so Kimi K2.5 is cheaper on this metric. Claude Haiku 4.5 is $5.00 and Kimi K2.5 is $2.25 per million tokens, so Kimi K2.5 is cheaper on this metric.
Which has a larger context window, Claude Haiku 4.5 or Kimi K2.5?+
Kimi K2.5 has the larger sourced context window. Claude Haiku 4.5 supports 200K and Kimi K2.5 supports 262K.
Which performs better in benchmarks, Claude Haiku 4.5 or Kimi K2.5?+
There is no overall benchmark winner: An overall winner requires at least two decisive benchmarks from at least two original publishers.
Can Claude Haiku 4.5 or Kimi K2.5 be self-hosted?+
Kimi K2.5 is the only model in this pair currently marked as self-hostable. Claude Haiku 4.5 is not marked open weight; Kimi K2.5 is open weight.
Can Claude Haiku 4.5 and Kimi K2.5 understand images?+
Claude Haiku 4.5 is documented with image input; Kimi K2.5 is documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, Claude Haiku 4.5 or Kimi K2.5?+
Neither has a larger sourced maximum output. Claude Haiku 4.5 is 64K and Kimi K2.5 is —.
Do Claude Haiku 4.5 and Kimi K2.5 support reasoning and tool use?+
Claude Haiku 4.5: reasoning, tool calling, and image input. Kimi K2.5: reasoning, tool calling, and image input. Feature support does not establish relative quality.
Which is available from more inference providers, Claude Haiku 4.5 or Kimi K2.5?+
Claude Haiku 4.5 has 2 sourced provider routes; Kimi K2.5 has 3, so Kimi K2.5 has broader tracked availability.
Which offers better value, Claude Haiku 4.5 or Kimi K2.5?+
There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.