vs
11:12
overall capability wins (1 ties across 24 scored). GPT-5.6 Sol leads overall.
- Reasoning2:4
- Coding6:3
- Arena1:0
- Agents1:3
- Tool use0:1
- Math1:0
- Composite indices2:0
- Pricing0:3
- Speed0:2
- Specs0:1
Verdict
Claude Fable 5 leads coding (widest gap: +15.4 on SWE-Pro); GPT-5.6 Sol edges reasoning & knowledge; Claude Fable 5 ranks higher on LMArena (+22 Elo).
- Claude Fable 5 leads coding (widest gap: +15.4 on SWE-Pro)
- GPT-5.6 Sol edges reasoning & knowledge
- Claude Fable 5 ranks higher on LMArena (+22 Elo)
- GPT-5.6 Sol offers larger context (1.1M vs 1M)
- GPT-5.6 Sol is ~2.5x cheaper on a blended token basis than Claude Fable 5
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Claude Fable 5
- $22.50
- GPT-5.6 Sol
- $9
GPT-5.6 Sol is about 2.5x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Anthropic
OpenAI
Family
Claude Fable / Mythos
GPT-5.x
License
Proprietary
Proprietary
Open weights
No
No
Release
Jun 9, 2026
Jul 9, 2026
Knowledge cutoff
-
2026-02-16
API / provider
Anthropic
OpenAI
Modalities
text, image → text
text, image → text
Specs
Context window
1M
1.1M
Max output
128K
128K
Parameters
—
—
Pricing
Input $/1M
$10
$4
Output $/1M
$50
$20
Blended $/1M (3∶1)
$20
$8
Speed
tok/s
58
96
TTFT (s)
0.7
0.45
Reasoning
MMLU-Pro
—
—
GPQA Diamond
92.6
94.1
Humanity's Last Exam
55.5
49.5
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
65.4
59.4
AA-LCR v1.1
82.3
84.0
CritPt
28.6
32.3
MMMU-Pro
—
83.4
IFBench
63.5
72.7
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
80.0
64.6
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
84.6
88.0
Aider Polyglot
—
—
Terminal-Bench 3
34.1
34.6
BigCodeBench
—
—
SciCode
61.0
57.1
CursorBench
70.5
67.2
SWE-Rebench
64.5
62.3
NL2Repo-Bench
—
—
DeepSWE
69.9
72.7
WebDev Arena
1,628
1,617
Terminal-Bench 4.0
44.6
37.3
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,506
1,484
| Metric | Claude Fable 5 | GPT-5.6 Sol | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Anthropic | OpenAI | — |
| Family | Claude Fable / Mythos | GPT-5.x | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Jun 9, 2026 | Jul 9, 2026 | — |
| Knowledge cutoff | - | 2026-02-16 | — |
| API / provider | Anthropic | OpenAI | — |
| Modalities | text, image → text | text, image → text | — |
| Specs | |||
| Context window | 1M | 1.1M | B +50K |
| Max output | 128K | 128K | tie |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $10 | $4 | B +$6 |
| Output $/1M | $50 | $20 | B +$30 |
| Blended $/1M (3∶1) | $20 | $8 | B +$12 |
| Speed | |||
| tok/s | 58 | 96 | B +38 |
| TTFT (s) | 0.7 | 0.45 | B +0.25 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 92.6 | 94.1 | B +1.5 pts |
| Humanity's Last Exam | 55.5 | 49.5 | A +6.0 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 65.4 | 59.4 | A +6.0 pts |
| AA-LCR v1.1 | 82.3 | 84.0 | B +1.7 pts |
| CritPt | 28.6 | 32.3 | B +3.7 pts |
| MMMU-Pro | — | 83.4 | — |
| IFBench | 63.5 | 72.7 | B +9.2 pts |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | 80.0 | 64.6 | A +15.4 pts |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 84.6 | 88.0 | B +3.4 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | 34.1 | 34.6 | B +0.5 pts |
| BigCodeBench | — | — | — |
| SciCode | 61.0 | 57.1 | A +3.9 pts |
| CursorBench | 70.5 | 67.2 | A +3.3 pts |
| SWE-Rebench | 64.5 | 62.3 | A +2.2 pts |
| NL2Repo-Bench | — | — | — |
| DeepSWE | 69.9 | 72.7 | B +2.8 pts |
| WebDev Arena | 1,628 | 1,617 | A +11 Elo |
| Terminal-Bench 4.0 | 44.6 | 37.3 | A +7.3 pts |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,506 | 1,484 | A +22 Elo |
Claude Fable 5
- Context
- 1M / 128K out
- Parameters
- —
- Price
- $10 / $50
- Speed
- 58 tok/s · 0.7s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Mythos-class Anthropic model optimized for long-horizon coding and writing; premium pricing relative to Opus 5.
GPT-5.6 Sol
- Context
- 1.1M / 128K out
- Parameters
- —
- Price
- $4 / $20
- Speed
- 96 tok/s · 0.45s TTFT
- Modalities
- text, image → text
- License
- Proprietary
OpenAI's July 2026 flagship reasoning tier, tuned for hard STEM and agentic work; ChatGPT's top API SKU in the 5.6 family. Promotional Standard rates are $4/$20 (through at least Nov 21, 2026). Fast mode (2× price) delivers up to 2.5× Standard speed, and the Cerebras-powered Ultrafast preview reaches up to 750 tok/s.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.