// head_to_head
Claude Opus 4.6 vs DeepSeek V4 Flash 0731
Published LiveBench scores across all seven categories, live list pricing, context windows, and the measured cost of a point of capability — for both models, side by side.
Effectively the same quality — DeepSeek V4 Flash 0731 is the cheaper way to get it.
The two are within 0.3 points overall, which is inside the range that effort settings alone move a LiveBench score, so treat them as quality-equivalent. DeepSeek V4 Flash 0731 lists 95× cheaper per blended million tokens. When quality ties, cost is the whole decision. DeepSeek V4 Flash 0731 also leads on measured cost per point of capability, at $0.0356 per point.
anthropic
Claude Opus 4.6
- Blended / 1M
- $10.00
- Context
- 1M
- Released
- Feb 4, 2026
- Overall score
- 74.5
deepseek
DeepSeek V4 Flash 0731
- Blended / 1M
- $0.105
- Context
- 1.3M
- Released
- Jul 31, 2026
- Overall score
- 74.2
Specs and pricing
| Metric | Claude Opus 4.6 | DeepSeek V4 Flash 0731 |
|---|---|---|
| LiveBench overall Mean of the seven LiveBench category scores, 0–100. Higher is better. | 74.5 | 74.2 |
| Cost per point Measured benchmark spend divided by overall score — dollars per point of capability. | $0.2101 | $0.0356win |
| Blended price / 1M 3:1 input:output mix, the usual shape of production traffic. | $10.00 | $0.105win |
| Input price / 1M | $5.00 | $0.080win |
| Output price / 1M | $25.00 | $0.180win |
| Cached input / 1M Price of an input token served from the prompt cache, where the provider publishes one. | $0.500 | $0.016win |
| Context window | 1M | 1.3Mwin |
| Max output tokens | 128K | 384Kwin |
Benchmarks by category
An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Claude Opus 4.6 on top, DeepSeek V4 Flash 0731 below, both out of 100.
What each one costs to run
Per-token prices are hard to feel. These are monthly list costs for both models across five workload shapes, using each provider's published cached-input rate where there is one.
| Workload | Claude Opus 4.6 | DeepSeek V4 Flash 0731 |
|---|---|---|
| Support chatbot 1.2K in / 400 out × 200K requests | $2876.00/mo | $28.99/mo |
| RAG assistant 8K in / 600 out × 100K requests | $3700.00/mo | $49.20/mo |
| Coding agent 40K in / 4K out × 20K requests | $3480.00/mo | $42.56/mo |
| Document extraction 20K in / 1.5K out × 50K requests | $6650.00/mo | $90.30/mo |
| Bulk classification 500 in / 20 out × 5M requests | $12,750/mo | $186.00/mo |
Which should you pick?
You are running this at volume
DeepSeek V4 Flash 0731
Lowest measured cost per point of capability at $0.0356 per point — the gap compounds with every request.
The workload is coding or agentic work
Claude Opus 4.6
Leads on agentic coding — 49.0 against 46.8.
You need to fit large documents in one call
DeepSeek V4 Flash 0731
Wider context window — 1.3M against 1M.
Claude Opus 4.6 vs DeepSeek V4 Flash 0731 FAQ
Which is better, Claude Opus 4.6 or DeepSeek V4 Flash 0731?
Effectively the same quality — DeepSeek V4 Flash 0731 is the cheaper way to get it. The two are within 0.3 points overall, which is inside the range that effort settings alone move a LiveBench score, so treat them as quality-equivalent. DeepSeek V4 Flash 0731 lists 95× cheaper per blended million tokens. When quality ties, cost is the whole decision. DeepSeek V4 Flash 0731 also leads on measured cost per point of capability, at $0.0356 per point.
Is Claude Opus 4.6 cheaper than DeepSeek V4 Flash 0731?
DeepSeek V4 Flash 0731 is cheaper. On a 3:1 input:output blend, Claude Opus 4.6 lists at $10.00 per million tokens and DeepSeek V4 Flash 0731 at $0.105 — DeepSeek V4 Flash 0731 is 95× cheaper. Input and output are priced separately — Claude Opus 4.6 charges $5.00 in and $25.00 out, DeepSeek V4 Flash 0731 charges $0.080 and $0.180 — so the model that looks cheaper flips depending on how output-heavy your workload is.
Claude Opus 4.6 vs DeepSeek V4 Flash 0731: which scores higher on benchmarks?
Claude Opus 4.6 scores 74.5 and DeepSeek V4 Flash 0731 scores 74.2 overall on LiveBench, the mean of its seven categories. That gap is inside the range that effort settings alone move a score, so treat them as equivalent on published quality. Category scores differ from the overall figure — a model can lead on reasoning and trail on coding, which the per-category table above breaks out.
Which gives better value for money, Claude Opus 4.6 or DeepSeek V4 Flash 0731?
DeepSeek V4 Flash 0731. Cost per point divides the measured dollars LiveBench spent running the benchmark by the score it earned, so it captures something token pricing misses: a reasoning model can emit many times more tokens than its per-token rate suggests. Claude Opus 4.6 works out at $0.2101 per point and DeepSeek V4 Flash 0731 at $0.0356.
Does Claude Opus 4.6 or DeepSeek V4 Flash 0731 have a bigger context window?
DeepSeek V4 Flash 0731 has the larger context window: 1M for Claude Opus 4.6 against 1.3M for DeepSeek V4 Flash 0731. Note that a window you can fill is not a window you should fill — retrieval quality usually degrades well before the limit, and you pay for every token you put in it.
Do Claude Opus 4.6 and DeepSeek V4 Flash 0731 support prompt caching?
Both publish a cached-input rate: $0.500 per million for Claude Opus 4.6 and $0.016 for DeepSeek V4 Flash 0731, against full input rates of $5.00 and $0.080. On a workload with a long stable prefix — a system prompt, a tool schema, a retrieved corpus — that changes the economics more than the headline price does.
Related comparisons
How these numbers are produced
- Price — provider list price from OpenRouter, refreshed every 15 minutes. “Blended” is a 3:1 input:output mix.
- Scores — LiveBench release 2026-06-25, using their own category map. Each model shows its strongest published run. A blank means “not evaluated”, never “bad”.
- Cost per point — the measured dollars LiveBench spent on the run, divided by the score it earned.
- “Win” — awarded only past a threshold: one full point on a benchmark score, 10% on a price, 25% on a context window. Anything tighter reports as a tie, because effort settings alone move a LiveBench score by more than that.
Published benchmarks rank models on someone else's tasks. Before committing, see LLM & agent evaluation for building an eval on your own.