Cheaper overall
GPT-5.4
72% lower blended rate
Higher benchmark scores
Claude Fable 5.1
leads on 20 of 20 figures
Bigger context window
GPT-5.4
1,050,000 vs 1,000,000 tokens
Where each one wins
Claude Fable 5.1
- Higher coding ability (Coding Index) 81.6 vs 71.1
- Higher graduate-level science (GPQA Diamond) 93.7 vs 92
- Higher expert-exam performance (Humanity's Last Exam) 59.1 vs 43.7
- Ahead on 17 more benchmark figures
GPT-5.4
- Cheaper input tokens $2.50 vs $10.00, 75% less
- Cheaper output tokens $15.00 vs $50.00, 70% less
- Cheaper on all four workloads
- Larger context window 1,050,000 tokens
- Offers Flex and Priority pricing not listed for Claude Fable 5.1
What four workloads cost
One run, and the same run a thousand times.
GPT-5.4 is cheaper on all four workloads, by much the same margin each time — about 3 times, from $0.0100 against $0.0350 on a chat turn to $2.07 against $4.15 on a long document.
| Workload | Claude Fable 5.1 | GPT-5.4 | ×1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0350 | $0.0100 | $35.00 / $10.00 | GPT-5.4 is 71% cheaper |
| RAG answer 10,000 in / 800 out | $0.1400 | $0.0370 | $140.00 / $37.00 | GPT-5.4 is 74% cheaper |
| Agent loop 32,000 in / 700 out × 20 calls | $1.25 | $0.4600 | $1,250.00 / $460.00 | GPT-5.4 is 63% cheaper |
| Long document 400,000 in / 3,000 out | $4.15 | $2.07 | $4,150.00 / $2,067.50 | GPT-5.4 is 50% cheaper |
Price per 1M tokens
Above 272,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point: GPT-5.4's input goes from $2.50 to $5.00. It does not change which of the two is cheaper.
| Context band | Price | Claude Fable 5.1 | GPT-5.4 | Difference |
|---|---|---|---|---|
| Prompts up to 272,000 tokens | Input | $10.00 | $2.50 | GPT-5.4 is 75% cheaper |
| Prompts up to 272,000 tokens | Cached input | $0.25 | $0.25 | Same |
| Prompts up to 272,000 tokens | Output | $50.00 | $15.00 | GPT-5.4 is 70% cheaper |
| Prompts over 272,000 tokens | Input | $10.00 | $5.00 | GPT-5.4 is 50% cheaper |
| Prompts over 272,000 tokens | Cached input | $0.25 | $0.50 | Claude Fable 5.1 is 50% cheaper |
| Prompts over 272,000 tokens | Output | $50.00 | $22.50 | GPT-5.4 is 55% cheaper |
Blended: Claude Fable 5.1 $20.00, GPT-5.4 $5.625 per 1M — one rate at 3:1 input to output, for comparing two models at a glance.
Claude Fable 5.1 priced by Anthropic, GPT-5.4 by OpenAI. Standard tier, pay-as-you-go.
Other pricing tiers
Both models offer Batch, worth up to 50% off their standard rates, and the cheaper model is the same one on every tier — so the tier you buy changes the bill rather than the choice. Only GPT-5.4 lists Flex and Priority, at up to 50% off its own standard rate; Claude Fable 5.1 offers none of them.
Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate — how we read tiers.
| Tier | Claude Fable 5.1 | GPT-5.4 | Cheaper |
|---|---|---|---|
| Batch | $5.00 in / $25.00 out 50% off Standard | $1.25 in / $7.50 out 50% off Standard | GPT-5.4 |
| Flex only on GPT-5.4 | Not offered | $1.25 in / $7.50 out 50% off Standard | — |
| Priority only on GPT-5.4 | Not offered | $5.00 in / $30.00 out | — |
Benchmarks
Claude Fable 5.1 leads on all 20 figures, furthest ahead on Fullstack Elo, by 309 points.
Coding ability
Coding Index — Coding ability across software-engineering evals
Claude Fable 5.1 ahead by 10.5
Graduate-level science
GPQA Diamond — Graduate-level questions in biology, chemistry & physics
Claude Fable 5.1 ahead by 1.7
Expert-exam performance
Humanity's Last Exam — Expert-level questions across many academic domains
Claude Fable 5.1 ahead by 15.4
Long-context reasoning
AA-LCR — Long-context reasoning across large inputs
Claude Fable 5.1 ahead by 3.3
Physics research reasoning
CritPt — Unpublished physics research reasoning problems
Claude Fable 5.1 ahead by 6.3
Real-world task performance
GDPval — Economically valuable, real-world knowledge work
Claude Fable 5.1 ahead by 25.1
Factual accuracy
Omniscience Accuracy — Breadth of factual knowledge across domains
Claude Fable 5.1 ahead by 16.4
Answer reliability
Omniscience Non-Hallucination — How reliably the model avoids fabricated answers
Claude Fable 5.1 ahead by 10
Fullstack Elo
Design Arena head-to-head rating in the fullstack category
Claude Fable 5.1 ahead by 309
Godotgamedev Elo
Design Arena head-to-head rating in the godotgamedev category
Claude Fable 5.1 ahead by 192
Mobileapps Elo
Design Arena head-to-head rating in the mobileapps category
Claude Fable 5.1 ahead by 153
Webapps Elo
Design Arena head-to-head rating in the webapps category
Claude Fable 5.1 ahead by 259
3d Elo
Design Arena head-to-head rating in the 3d category
Claude Fable 5.1 ahead by 289
Asciiart Elo
Design Arena head-to-head rating in the asciiart category
Claude Fable 5.1 ahead by 124
Codecategories Elo
Design Arena head-to-head rating in the codecategories category
Claude Fable 5.1 ahead by 123
Dataviz Elo
Design Arena head-to-head rating in the dataviz category
Claude Fable 5.1 ahead by 112
Gamedev Elo
Design Arena head-to-head rating in the gamedev category
Claude Fable 5.1 ahead by 141
Svg Elo
Design Arena head-to-head rating in the svg category
Claude Fable 5.1 ahead by 113
Uicomponent Elo
Design Arena head-to-head rating in the uicomponent category
Claude Fable 5.1 ahead by 98
Website Elo
Design Arena head-to-head rating in the website category
Claude Fable 5.1 ahead by 91
Each figure is that model's best published run, with the effort level named beside it — where these scores come from.
Specifications
GPT-5.4 holds 50,000 more input tokens in one request.
| Specification | Claude Fable 5.1 | GPT-5.4 |
|---|---|---|
| Context window | 1,000,000 tokens | 1,050,000 tokens |
| Max output | 128,000 tokens | 128,000 tokens |
| Takes in | Text, image, file | Text, image, file |
| Puts out | Text | Text |
| Knowledge cutoff | 30 Jun 2026 | 31 Aug 2025 |
| Released | 1 Sep 2026 | 5 Mar 2026 |
| Status | Active | Active |
| Sold by | Anthropic | OpenAI, Perplexity |
FAQs
- Is Claude Fable 5.1 cheaper than GPT-5.4?
- No — the other way round. A 1,000-token prompt with a 500-token reply costs $0.0350 on Claude Fable 5.1 and $0.0100 on GPT-5.4, and the same model is cheaper on every workload on this page.
- How much do Claude Fable 5.1 and GPT-5.4 cost per 1M tokens?
- Claude Fable 5.1 costs $10.00 for input and $50.00 for output. GPT-5.4 costs $2.50 and $15.00. Standard pay-as-you-go rates, as of 1 Sep 2026.
- Which is better, Claude Fable 5.1 or GPT-5.4?
- On published benchmarks, Claude Fable 5.1. It leads on 20 of the 20 figures both models report, including coding ability (Coding Index), where it scores 81.6 against 71.1.
- Is Batch pricing cheaper for Claude Fable 5.1 or GPT-5.4?
- Batch input costs $5.00 on Claude Fable 5.1 and $1.25 on GPT-5.4, 50% below the standard rate on both.
- Does Claude Fable 5.1 or GPT-5.4 have a bigger context window?
- GPT-5.4, at 1,050,000 tokens against 1,000,000. That is 50,000 more input tokens in a single request.