ZMIME
Comparison · 3 models · Updated Nov 18, 2026

GPT-5.4 Pro vs Kimi K3 vs Claude Opus 4.8

Too close to call on our weighted score (Kimi K3 64, Claude Opus 4.8 62, GPT-5.4 Pro 59). The right pick depends on what you value most.

Price in rating: Included · Exclude price

  1. OpenAI

    GPT-5.4 Pro

    Released Mar 5, 2026

    59/100
    • ECI160.3
    • Price$30.00 / $180.00
    • Context1.05M
  2. Moonshot AI

    Kimi K3

    Released Jul 16, 2026

    64/100
    • ECI158.6
    • Price$3.00 / $15.00
    • Context1.05M
  3. Anthropic

    Claude Opus 4.8

    Released May 28, 2026

    62/100
    • ECI159.9
    • Price$5.00 / $25.00
    • Context1M
01 — Verdict

Too close to call

It is close. Our weighted score puts them within 2 points (Kimi K3 64/100, Claude Opus 4.8 62/100, GPT-5.4 Pro 59/100), so choose by what matters most for your work: GPT-5.4 Pro for raw capability and Kimi K3 on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

  • CapabilityGPT-5.4 ProCapabilities Index (ECI): GPT-5.4 Pro 160.3 · Claude Opus 4.8 159.9 · Kimi K3 158.6
  • Lowest priceKimi K3Kimi K3 $6.00 · Claude Opus 4.8 $10.00 · GPT-5.4 Pro $67.50 per 1M tokens (3:1 blend)
  • Longest contextGPT-5.4 Pro and Kimi K3GPT-5.4 Pro 1,050,000 · Kimi K3 1,048,576 · Claude Opus 4.8 1,000,000 tokens
  • Widest inputsKimi K3 and Claude Opus 4.8GPT-5.4 Pro: Text, Images · Kimi K3: Text, Images, Video · Claude Opus 4.8: Text, Images, PDFs
  • Self-hostingKimi K3Publishes downloadable weights
How the score is built
MeasureWeightGPT-5.4 ProKimi K3Claude Opus 4.8
CapabilityCapabilities Index (ECI)50%878587
Price25%0122
Inputs & features15%608080
Context window10%656564
Overall100%59/10064/10062/100
02 — Side by side

Every spec in one table

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

GPT-5.4 Pro vs Kimi K3 vs Claude Opus 4.8 specifications side by side
SpecificationGPT-5.4 ProOpenAIKimi K3Moonshot AIClaude Opus 4.8Anthropic
Capability
Capabilities Index (ECI)160.3 (best)158.6159.9
ECI rank#19 of 158 (best)#25 of 158#22 of 158
GPQA DiamondGraduate-level science questions94.6% (best)93.1%91.0%
FrontierMath Tiers 1–3Research-level mathematics82.5% (best)72.2%80.0%
OTIS Mock AIME 2024–2025Competition mathematics—97.2%98.3% (best)
SimpleQA VerifiedShort factual questions46.3%50.6%53.0% (best)
Price per million tokens
Input$30.00$3.00 (best)$5.00
Output$180.00$15.00 (best)$25.00
Cached input—$0.30 (best)$0.50
Blended (3:1)$67.50$6.00 (best)$10.00
Long-context rateOver 272K: $60.00 / $270.00Same rateSame rate
Price sourceOfficial OpenAI APIOfficial Moonshot AI APIOfficial Anthropic API
Limits
Context window1,050,000 tokens (best)1,048,576 tokens1,000,000 tokens
Max output128,000 tokens1,048,576 tokens (best)128,000 tokens
Inputs and features
TextYesYesYes
ImagesYesYesYes
PDFsNoNoYes
AudioNoNoNo
VideoNoYesNo
ReasoningYesmedium · high · xhighYeslow · high · maxYeslow · medium · high · xhigh · max
Tool callingYesYesYes
Structured outputNoYesYes
Availability
WeightsProprietaryOpenProprietary
API model IDgpt-5.4-prokimi-k3claude-opus-4-8
API providers2294 (best)57
ReleasedMar 5, 2026Jul 16, 2026May 28, 2026
Knowledge cutoffAug 31, 2025—Jan 2026
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.

  • GPT-5.4 Pro$660.00
  • Kimi K3$60.00
  • Claude Opus 4.8$100.00
04 — Questions

Which should you choose?

Which is better: GPT-5.4 Pro, Kimi K3 or Claude Opus 4.8?

It is close. Our weighted score puts them within 2 points (Kimi K3 64/100, Claude Opus 4.8 62/100, GPT-5.4 Pro 59/100), so choose by what matters most for your work: GPT-5.4 Pro for raw capability and Kimi K3 on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Which is cheaper, GPT-5.4 Pro, Kimi K3 or Claude Opus 4.8?

Kimi K3 is cheaper at $3.00 input / $15.00 output per million tokens (official Moonshot AI API price). Claude Opus 4.8 costs $5.00 input / $25.00 output per million tokens (official Anthropic API price); GPT-5.4 Pro costs $30.00 input / $180.00 output per million tokens (official OpenAI API price). At a typical mix of three input tokens to one output token, that is $6.00 per million tokens for Kimi K3 versus $10.00 for Claude Opus 4.8 (1.7× as much) and $67.50 for GPT-5.4 Pro (11× as much).

Which scores higher on benchmarks?

GPT-5.4 Pro scores higher on the Capabilities Index (ECI): GPT-5.4 Pro 160.3 (#19 of 158), Claude Opus 4.8 159.9 (#22 of 158) and Kimi K3 158.6 (#25 of 158). The confidence ranges of the top two overlap (157.9–163.0 vs 157.8–162.2), so treat the gap as small. On individual benchmarks: GPQA Diamond — GPT-5.4 Pro 94.6%, Kimi K3 93.1%, Claude Opus 4.8 91.0%; FrontierMath Tiers 1–3 — GPT-5.4 Pro 82.5%, Claude Opus 4.8 80.0%, Kimi K3 72.2%; SimpleQA Verified — Claude Opus 4.8 53.0%, Kimi K3 50.6%, GPT-5.4 Pro 46.3%.

Which is better for coding?

There are no published SWE-bench Verified results for GPT-5.4 Pro, Kimi K3 and Claude Opus 4.8 yet, so there is no like-for-like coding score. On overall capability, GPT-5.4 Pro 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?

GPT-5.4 Pro and Kimi K3 have the largest context windows (1,050,000 and 1,048,576 tokens), against 1,000,000 for Claude Opus 4.8. Maximum output per response: GPT-5.4 Pro up to 128,000, Kimi K3 up to 1,048,576, Claude Opus 4.8 up to 128,000 tokens.

Which can read images, PDFs, audio or video?

GPT-5.4 Pro accepts text and images; Kimi K3 accepts text, images and video; Claude Opus 4.8 accepts text, images and PDFs. Kimi K3 handles the widest range of inputs.

Are any of these open source?

Kimi K3 publishes its weights and can be self-hosted; GPT-5.4 Pro and Claude Opus 4.8 is proprietary.

Which is newer?

Kimi K3 is the newest, released Jul 16, 2026. Claude Opus 4.8 came out May 28, 2026; GPT-5.4 Pro came out Mar 5, 2026. Knowledge cutoff: GPT-5.4 Pro Aug 31, 2025, Claude Opus 4.8 Jan 2026.

How do you decide the winner?

This rating uses capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, 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.