ZMIME
Comparison · 3 models · Updated Nov 18, 2026

Claude Haiku 4.5 vs Kimi K2 Instruct vs Qwen3 Max

Claude Haiku 4.5 comes out ahead, 68 to 57 and 55 on our weighted score, though Kimi K2 Instruct is 2× cheaper per token.

Price in rating: Excluded · Include price

  1. Our pick

    Anthropic

    Claude Haiku 4.5

    Released Oct 15, 2025

    68/100
    • ECI142.8
    • Price$1.00 / $5.00
    • Context200K
  2. Moonshot AI

    Kimi K2 Instruct

    Released Jul 11, 2025

    55/100
    • ECI141.3
    • Price$0.57 / $2.30
    • Context131K
  3. Alibaba (Qwen)

    Qwen3 Max

    Released Sep 23, 2025Deprecated

    57/100
    • ECI142.4
    • Price$1.20 / $6.00
    • Context262K
01 — Verdict

Claude Haiku 4.5 is our pick

Claude Haiku 4.5 is the better all-round choice, scoring 68/100 against Qwen3 Max (57) and Kimi K2 Instruct (55). It leads on inputs & features. Qwen3 Max wins on context window. The score weighs capability 67%, inputs & features 20%, context window 13%.

  • CapabilityClaude Haiku 4.5Capabilities Index (ECI): Claude Haiku 4.5 142.8 · Qwen3 Max 142.4 · Kimi K2 Instruct 141.3
  • Lowest priceKimi K2 InstructKimi K2 Instruct $1.00 · Claude Haiku 4.5 $2.00 · Qwen3 Max $2.40 per 1M tokens (3:1 blend)
  • Longest contextQwen3 MaxQwen3 Max 262,144 · Claude Haiku 4.5 200,000 · Kimi K2 Instruct 131,072 tokens
  • Widest inputsClaude Haiku 4.5Claude Haiku 4.5: Text, Images, PDFs · Kimi K2 Instruct: Text · Qwen3 Max: Text
  • Self-hostingKimi K2 InstructPublishes downloadable weights (Modified MIT)
How the score is built
MeasureWeightClaude Haiku 4.5Kimi K2 InstructQwen3 Max
CapabilityCapabilities Index (ECI)67%706970
Inputs & features20%802525
Context window13%393243
Overall100%68/10055/10057/100
02 — Side by side

Every spec in one table

Highlighted cells lead their row. Dashes mean the data is not published.

Claude Haiku 4.5 vs Kimi K2 Instruct vs Qwen3 Max specifications side by side
SpecificationClaude Haiku 4.5AnthropicKimi K2 InstructMoonshot AIQwen3 MaxAlibaba (Qwen)
Capability
Capabilities Index (ECI)142.8 (best)141.3142.4
ECI rank#102 of 158 (best)#108 of 158#103 of 158
GPQA DiamondGraduate-level science questions71.2%—72.6% (best)
FrontierMath Tiers 1–3Research-level mathematics——19.0%
OTIS Mock AIME 2024–2025Competition mathematics66.7%—73.3% (best)
SimpleQA VerifiedShort factual questions13.2%—48.8% (best)
Price per million tokens
Input$1.00$0.57 (best)$1.20
Output$5.00$2.30 (best)$6.00
Cached input$0.10——
Blended (3:1)$2.00$1.00 (best)$2.40
Long-context rateSame rateSame rateSame rate
Price sourceOfficial Anthropic APIMedian of 11 providersOfficial Alibaba API
Limits
Context window200,000 tokens131,072 tokens262,144 tokens (best)
Max output64,000 tokens32,768 tokens65,536 tokens (best)
Inputs and features
TextYesYesYes
ImagesYesNoNo
PDFsYesNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningYesNoNo
Tool callingYesYesYes
Structured outputYesNoNo
Availability
WeightsProprietaryOpenModified MITProprietary
API model IDclaude-haiku-4-5—qwen3-max
API providers57 (best)1113
ReleasedOct 15, 2025Jul 11, 2025Sep 23, 2025
Knowledge cutoffFeb 28, 2025—Apr 2025
03 — Cost

What would a month cost?

Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.

  • Claude Haiku 4.5$20.00
  • Kimi K2 Instruct$10.30
  • Qwen3 Max$24.00
04 — Questions

Which should you choose?

Which is better: Claude Haiku 4.5, Kimi K2 Instruct or Qwen3 Max?

Claude Haiku 4.5 is the better all-round choice, scoring 68/100 against Qwen3 Max (57) and Kimi K2 Instruct (55). It leads on inputs & features. Qwen3 Max wins on context window. The score weighs capability 67%, inputs & features 20%, context window 13%.

Which is cheaper, Claude Haiku 4.5, Kimi K2 Instruct or Qwen3 Max?

Kimi K2 Instruct is cheaper at $0.57 input / $2.30 output per million tokens (median across 11 API providers). Claude Haiku 4.5 costs $1.00 input / $5.00 output per million tokens (official Anthropic API price); Qwen3 Max costs $1.20 input / $6.00 output per million tokens (official Alibaba API price). At a typical mix of three input tokens to one output token, that is $1.00 per million tokens for Kimi K2 Instruct versus $2.00 for Claude Haiku 4.5 (2× as much) and $2.40 for Qwen3 Max (2.4× as much).

Which scores higher on benchmarks?

Claude Haiku 4.5 scores higher on the Capabilities Index (ECI): Claude Haiku 4.5 142.8 (#102 of 158), Qwen3 Max 142.4 (#103 of 158) and Kimi K2 Instruct 141.3 (#108 of 158). The confidence ranges of the top two overlap (141.2–144.3 vs 140.3–144.5), so treat the gap as small.

Which is better for coding?

There are no published SWE-bench Verified results for Claude Haiku 4.5, Kimi K2 Instruct and Qwen3 Max yet, so there is no like-for-like coding score. On overall capability, Claude Haiku 4.5 leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.

Which has the bigger context window?

Qwen3 Max has the largest context window at 262,144 tokens, against 200,000 for Claude Haiku 4.5 and 131,072 for Kimi K2 Instruct. Maximum output per response: Claude Haiku 4.5 up to 64,000, Kimi K2 Instruct up to 32,768, Qwen3 Max up to 65,536 tokens.

Which can read images, PDFs, audio or video?

Claude Haiku 4.5 accepts text, images and PDFs; Kimi K2 Instruct accepts text; Qwen3 Max accepts text. Claude Haiku 4.5 handles the widest range of inputs.

Are any of these open source?

Kimi K2 Instruct publishes its weights (Modified MIT) and can be self-hosted; Claude Haiku 4.5 and Qwen3 Max is proprietary.

Which is newer?

Claude Haiku 4.5 is the newest, released Oct 15, 2025. Qwen3 Max came out Sep 23, 2025; Kimi K2 Instruct came out Jul 11, 2025. Knowledge cutoff: Claude Haiku 4.5 Feb 28, 2025, Qwen3 Max Apr 2025.

How do you decide the winner?

This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 18, 2026.