ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

Claude Opus 4.8vsClaude Opus 5

Claude Opus 5 leads 6 of 6 shared benchmarks. Both list the same combined token price. Both offer a 1M-token context window.

Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.

Shared benchmarks
0 – 6
Claude Opus 5 leads
Cheaper per token
Tie
list price, input + output
Larger context
Tie
both 1M tokens
Provider uptime (30d)
100%
Anthropic
Claude Opus 4.8
Anthropic · released 2026-05-28
textvision
Context
1M
Max out
128K
Input /1M
$5
Output /1M
$25
Cached /1M
$0.5
Scores
17 · 13 core
Claude Opus 5
Anthropic · released 2026-07-24
textvision
Context
1M
Max out
128K
Input /1M
$5
Output /1M
$25
Cached /1M
$0.5
Scores
9 · 6 core
Quality

Benchmark matrix

BenchmarkClaude Opus 4.8Claude Opus 5ΔEdge
Reported by both · 6
GPQA Diamond93.6% 93.9% epoch-0.3 ptClaude Opus 5
Humanity's Last Exam (no tools)49.8% 56.6% -6.8 ptClaude Opus 5
FrontierMath Tier 4 v2 (Epoch AI run)56.1% epoch73.2% epoch-17.1 ptClaude Opus 5
FrontierMath Tiers 1-3 v2 (Epoch AI run)80% epoch85.6% epoch-5.6 ptClaude Opus 5
OTIS Mock AIME 2024-2025 (Epoch AI run)98.3% epoch98.9% epoch-0.6 ptClaude Opus 5
SimpleQA Verified53% epoch59.9% epoch-6.9 ptClaude Opus 5
Only Claude Opus 4.8 reports · 10
SWE-bench Verified88.6% not reported
AIME 2026100% matharenanot reported
LMArena Elo1452.7 not reported
BrowseComp84.3% not reported
Humanity's Last Exam (with tools)57.9% not reported
OSWorld-Verified83.4% not reported
SWE-bench Multilingual84.4% not reported
SWE-bench Multimodal38.4% not reported
SWE-Bench Pro69.2% not reported
Terminal-Bench 2.174.6% not reported
Only Claude Opus 5 reports · 3
CursorBench 3.2.0not reported70%
GDPval-AA v2not reported1824
Terminal-Bench Science 0.1not reported29%
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Claude Opus 4.8 minus Claude Opus 5 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing

Side by side

SpecClaude Opus 4.8Claude Opus 5Edge
Context window1M tokens1M tokensTie
Max output128K tokens128K tokensTie
Input price / 1M$5 $5 Tie
Output price / 1M$25 $25 Tie
Cached input / 1M$0.5 $0.5 Tie
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
$30$30Tie
Modalitiestext · visiontext · visionTie
Released2026-05-282026-07-24
Cited benchmark scores179
Reliability

Anthropic status

All providers
More matchups

Claude Opus 4.8 vs …

More matchups

Claude Opus 5 vs …

Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run Claude Opus 4.8 and Claude Opus 5 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.