ZMIME
Comparison · 3 models · Updated Nov 22, 2026

GPT-5.4 Pro vs GPT-5.2 Pro vs Claude Fable 5

Claude Fable 5 comes out ahead, 64 to 59 and 56 on our weighted score, and it is the cheaper option too.Claude Fable 5 comes out ahead, 85 to 79 and 75 on our weighted score, and it is the cheaper option too.

  1. OpenAI

    GPT-5.4 Pro

    Released Mar 5, 2026

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

    GPT-5.2 Pro

    Released Dec 11, 2025

    56/100
    • ECI157.4
    • Price$21.00 / $168.00
    • Context400K
  3. Our pick

    Anthropic

    Claude Fable 5

    Released Jun 9, 2026

    64/100
    • ECI164.2
    • Price$10.00 / $50.00
    • Context1M
  4. OpenAI

    GPT-5.4 Pro

    Released Mar 5, 2026

    79/100
    • ECI160.3
    • Price$30.00 / $180.00
    • Context1.05M
  5. OpenAI

    GPT-5.2 Pro

    Released Dec 11, 2025

    75/100
    • ECI157.4
    • Price$21.00 / $168.00
    • Context400K
  6. Our pick

    Anthropic

    Claude Fable 5

    Released Jun 9, 2026

    85/100
    • ECI164.2
    • Price$10.00 / $50.00
    • Context1M
01 — Verdict

Claude Fable 5 is our pick

Claude Fable 5 is the better all-round choice, scoring 64/100 against GPT-5.4 Pro (59) and GPT-5.2 Pro (56). It leads on capability and inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.Claude Fable 5 is the better all-round choice, scoring 85/100 against GPT-5.4 Pro (79) and GPT-5.2 Pro (75). It leads on capability and inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.

  • CapabilityClaude Fable 5Capabilities Index (ECI): Claude Fable 5 164.2 · GPT-5.4 Pro 160.3 · GPT-5.2 Pro 157.4
  • Lowest priceClaude Fable 5Claude Fable 5 $20.00 · GPT-5.2 Pro $57.75 · GPT-5.4 Pro $67.50 per 1M tokens (3:1 blend)
  • Longest contextGPT-5.4 ProGPT-5.4 Pro 1,050,000 · Claude Fable 5 1,000,000 · GPT-5.2 Pro 400,000 tokens
  • Widest inputsClaude Fable 5GPT-5.4 Pro: Text, Images · GPT-5.2 Pro: Text, Images · Claude Fable 5: Text, Images, PDFs
  • Self-hostingNo confirmed open weightsAll three are available only through APIs
How the score is built
MeasureWeightGPT-5.4 ProGPT-5.2 ProClaude Fable 5
CapabilityCapabilities Index (ECI)50%888591
Price25%000
Inputs & features15%606080
Context window10%655064
Overall100%59/10056/10064/100
How the score is built
MeasureWeightGPT-5.4 ProGPT-5.2 ProClaude Fable 5
CapabilityCapabilities Index (ECI)67%888591
Inputs & features20%606080
Context window13%655064
Overall100%79/10075/10085/100
02 — Side by side

Every spec in one table

Highlighted cells lead their row. Unavailable specifications are marked “Not published”. Publisher results include their evaluation settings and source.

GPT-5.4 Pro vs GPT-5.2 Pro vs Claude Fable 5 specifications side by side
SpecificationGPT-5.4 ProOpenAIGPT-5.2 ProOpenAIClaude Fable 5Anthropic
Capability
Capabilities Index (ECI)160.3157.4164.2 (best)
ECI rank#21 of 159#33 of 159#13 of 159 (best)
ECI confidence interval157.9–163.0155.1–160.0161.8–166.9
ECI evaluation dateMar 5, 2026Dec 11, 2025Jun 9, 2026
Documented capabilitiesMore exact GPT-5.4 tier for demanding professional reasoning and agent tasksHigher-accuracy GPT-5.2 variant for tougher reasoning and review workflowsClaude model for creative writing, analysis, and controlled agent workflows
Model parametersNot publishedNot publishedNot published
Output modalitiesTextTextText
GPQA DiamondGraduate-level science questions94.6%xhigh (best)Not published93.9%max
FrontierMath Tiers 1–3Research-level mathematics82.5%xhigh74.0%xhigh87.0%max (best)
OTIS Mock AIME 2024–2025Competition mathematicsNot publishedNot published100%high
SimpleQA VerifiedShort factual questions46.3%xhighNot published70.7%xhigh (best)
Publisher-reported benchmarks
GPQA DiamondPublisher-reported result94.4accuracyReported resultApr 23, 2026Not publishedNot published
Humanity's Last Examno tools42.7accuracyReported resultApr 23, 2026Not published59accuracyReported resultJun 9, 2026
Humanity's Last Examwith tools58.7accuracyReported resultApr 23, 2026Not published64.5accuracyReported resultJun 9, 2026
BrowseCompPublisher-reported result89.3accuracyReported resultApr 23, 2026Not publishedNot published
GDPvalPublisher-reported result82wins or tiesReported resultApr 23, 2026Not publishedNot published
FrontierMath · Tier 1-3Publisher-reported result50accuracyReported resultApr 23, 2026Not publishedNot published
FrontierMath · Tier 4Publisher-reported result38accuracyReported resultApr 23, 2026Not publishedNot published
ARC-AGI-1Verified94.5accuracyReported resultApr 23, 2026Not publishedNot published
ARC-AGI-2Verified83.3accuracyReported resultApr 23, 2026Not publishedNot published
FinanceAgent 1.1Publisher-reported result61.5accuracyReported resultApr 23, 2026Not publishedNot published
GeneBenchPublisher-reported result25.6accuracyReported resultApr 23, 2026Not publishedNot published
SWE-Bench ProPublisher-reported resultNot publishedNot published80.3resolve rateReported resultJun 9, 2026
SWE-Bench VerifiedPublisher-reported resultNot publishedNot published95resolvedReported result
Terminal-Bench 2.1Publisher-reported resultNot publishedNot published88success rateReported resultJun 9, 2026
OSWorld-VerifiedPublisher-reported resultNot publishedNot published85success rateReported resultJun 9, 2026
FrontierCode · Diamondhigh effortNot publishedNot published29.3pass rateReported resultJun 9, 2026
GDPval-AAPublisher-reported resultNot publishedNot published1932EloReported resultJun 9, 2026
AutomationBenchPublisher-reported resultNot publishedNot published17.4success rateReported resultJun 9, 2026
Price per million tokens
Input$30.00$21.00$10.00 (best)
Output$180.00$168.00$50.00 (best)
Cached inputNot publishedNot published$1.00
Blended (3:1)$67.50$57.75$20.00 (best)
Long-context rateOver 272K: $60.00 / $270.00Same rateSame rate
Price sourceOfficial OpenAI APIOfficial OpenAI APIOfficial Anthropic API
Limits
Context window1,050,000 tokens (best)400,000 tokens1,000,000 tokens
Max input922,000 tokens (best)272,000 tokensNot published
Max output128,000 tokens128,000 tokens128,000 tokens
Inputs and features
TextYesYesYes
ImagesYesYesYes
PDFsNoNoYes
AudioNoNoNo
VideoNoNoNo
ReasoningYesmedium · high · xhighYesmedium · high · xhighYeslow · medium · high · xhigh · max
Tool callingYesYesYes
Structured outputNoNoYes
File attachmentsYesYesYes
Temperature controlNoNoNo
Reasoning controleffortefforteffort
Availability
WeightsProprietaryProprietaryProprietary
API model IDgpt-5.4-progpt-5.2-proclaude-fable-5
API providers221355 (best)
ReleasedMar 5, 2026Dec 11, 2025Jun 9, 2026
Knowledge cutoffAug 31, 2025Aug 31, 2025Jan 31, 2026
Publisher documentationPublisher documentationPublisher documentationPublisher documentation
Source metadata updatedMar 5, 2026Dec 11, 2025Jun 9, 2026
Data sourcesmodels.dev · Epoch AImodels.dev · Epoch AImodels.dev · Epoch AI
Catalog checkedNov 22, 2026Nov 22, 2026Nov 22, 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
  • GPT-5.2 Pro$546.00
  • Claude Fable 5$200.00
04 — Questions

Which should you choose?

Which is better: GPT-5.4 Pro, GPT-5.2 Pro or Claude Fable 5?

Claude Fable 5 is the better all-round choice, scoring 64/100 against GPT-5.4 Pro (59) and GPT-5.2 Pro (56). It leads on capability and inputs & features. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.Claude Fable 5 is the better all-round choice, scoring 85/100 against GPT-5.4 Pro (79) and GPT-5.2 Pro (75). It leads on capability and inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.

Which is cheaper, GPT-5.4 Pro, GPT-5.2 Pro or Claude Fable 5?

Claude Fable 5 is cheaper at $10.00 input / $50.00 output per million tokens (official Anthropic API price). GPT-5.2 Pro costs $21.00 input / $168.00 output per million tokens (official OpenAI 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 $20.00 per million tokens for Claude Fable 5 versus $57.75 for GPT-5.2 Pro (2.9× as much) and $67.50 for GPT-5.4 Pro (3.4× as much).

Which scores higher on benchmarks?

Claude Fable 5 scores higher on the Capabilities Index (ECI): Claude Fable 5 164.2 (#13 of 159), GPT-5.4 Pro 160.3 (#21 of 159) and GPT-5.2 Pro 157.4 (#33 of 159). The confidence ranges of the top two overlap (161.8–166.9 vs 157.9–163.0), so treat the gap as small. On individual benchmarks: FrontierMath Tiers 1–3 — Claude Fable 5 87.0%, GPT-5.4 Pro 82.5%, GPT-5.2 Pro 74.0%.

Which is better for coding?

There are no published SWE-bench Verified results for GPT-5.4 Pro, GPT-5.2 Pro and Claude Fable 5 yet, so there is no like-for-like coding score. On overall capability, Claude Fable 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?

GPT-5.4 Pro has the largest context window at 1,050,000 tokens, against 1,000,000 for Claude Fable 5 and 400,000 for GPT-5.2 Pro. Maximum output per response: GPT-5.4 Pro up to 128,000, GPT-5.2 Pro up to 128,000, Claude Fable 5 up to 128,000 tokens.

Which can read images, PDFs, audio or video?

GPT-5.4 Pro accepts text and images; GPT-5.2 Pro accepts text and images; Claude Fable 5 accepts text, images and PDFs. Claude Fable 5 handles the widest range of inputs.

Are any of these open source?

No. GPT-5.4 Pro, GPT-5.2 Pro and Claude Fable 5 are proprietary and only available through APIs and apps.

Which is newer?

Claude Fable 5 is the newest, released Jun 9, 2026. GPT-5.4 Pro came out Mar 5, 2026; GPT-5.2 Pro came out Dec 11, 2025. Knowledge cutoff: GPT-5.4 Pro Aug 31, 2025, GPT-5.2 Pro Aug 31, 2025, Claude Fable 5 Jan 31, 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 22, 2026.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 22, 2026.