HEAD TO HEAD
Make the
right call.
Pick two to four models. Compare category evidence, price, context, speed and capabilities side by side.
01 / SCORECARD
Coding focus · side by side
| Metric | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| Quantum Score | Insufficient verified data73% coverage | 51QUANTUM SCORE |
| Confidence / coverage | Limited evidence · 73% · 2 independent domains | Moderate confidence · 73% · 4 independent domains |
| Context window | 1,000,000 tokens | 1,050,000 tokens |
| Modalities / tools | text · tool use | text, image · tool use, vision |
| Coding | 33 | 61 +28 vs first |
| Reasoning | 95 | 51 -44 vs first |
| Math | 79 | 67 -12 vs first |
REVAL™ⓘREVAL™ Research Evaluation Index — our proprietary 0–100 benchmark measuring real-world AI research performance. Methodology → | Not tested | Not tested |
| Agents | 40 | 20 -20 vs first |
| Multimodal | No data | No data |
| Speed | Not measured | Not measured |
| Value | 52 | 48 -4 vs first |
| Input / 1M | $1.32 | $0.20 |
| Output / 1M | $3.96 | $1.20 |
| Measured output speed | Not measured | Not measured |
THE DECISION
What this comparison means
Quality for your coding focus
3 current independent evaluations use matching versions, methods, and test settings across every selected model. These show performance on the named tests; they do not establish a universal best model.
DeepSeek V4 Pro leads on Terminal-Bench 4.0 Vals mini-SWE-agent avg@3 fallbacks failed
4.0 · Vals mini-SWE-agent avg@3; Anthropic fallbacks treated as failures
- DeepSeek V4 Pro: 14.141 % · max effort · Evidence ↗ · verified 2026-10-01
- GPT-5.6 Luna: 11.616 % · max effort · Evidence ↗ · verified 2026-10-01
GPT-5.6 Luna leads on LiveBench code_completion
2026-06-25 · Public table; published model effort variant · LiveBench public table 2026-06-25
- DeepSeek V4 Pro: 78.261 % · deepseek-v4-pro-0813 effort · Evidence ↗ · verified 2026-09-30
- GPT-5.6 Luna: 86.957 % · gpt-5.6-luna-max effort · Evidence ↗ · verified 2026-10-01
GPT-5.6 Luna leads on LiveBench code_generation
2026-06-25 · Public table; published model effort variant · LiveBench public table 2026-06-25
- DeepSeek V4 Pro: 76.056 % · deepseek-v4-pro-0813 effort · Evidence ↗ · verified 2026-09-30
- GPT-5.6 Luna: 78.873 % · gpt-5.6-luna-max effort · Evidence ↗ · verified 2026-10-01
Different benchmarks test different tasks. Effort levels are shown because they can affect results and cost. Verification dates indicate when we checked a source; they are not necessarily evaluation dates.
Your estimated monthly API cost
Enter usage in millions of tokens. Estimates use the listed standard input and output prices in USD.
DeepSeek V4 Pro
$21.12Pricing source ↗ · verified 2026-09-29
Price scope
API ID deepseek-v4-pro. Peak cache-miss input rate; off-peak and cache-hit rates differ
GPT-5.6 Luna
$4.40Lowest estimated cost among these models
Pricing source ↗ · verified 2026-10-01
Price scope
Standard short-context token rate; long-context and service tiers differ
This estimates API token charges. Subscription plans, caching, batch discounts, tools, taxes, and differences in tokens used per task can change the bill. Cost alone does not establish equivalent quality.
Tradeoffs and evidence gaps
DeepSeek V4 Pro ↗
Listed capabilities: tool use.
Independent benchmark records cover coding, agents, math, reasoning. Coverage does not guarantee comparable tests for every selected model.
GPT-5.6 Luna ↗
Listed capabilities: tool use, vision.
Independent benchmark records cover reasoning, coding, agents, math. Coverage does not guarantee comparable tests for every selected model.
How we assess scores and missing evidence →
Keep this comparison
Create a free account to save these models and get alerts when their verified evidence or prices change.
Sign in to save ↗02 / EVIDENCE
Published benchmark evidence
| Benchmark | DeepSeek V4 Pro | GPT-5.6 Luna |
|---|---|---|
| Terminal-Bench 4.0 Vals mini-SWE-agent avg@3 fallbacks failed · 4.0 · Vals mini-SWE-agent avg@3; Anthropic fallbacks treated as failures | 14.141 % independent lab ↗ | 11.616 % independent lab ↗ |
| LiveBench code_completion · 2026-06-25 · Public table; published model effort variant | 78.261 % benchmark maintainer ↗ | 86.957 % benchmark maintainer ↗ |
| LiveBench code_generation · 2026-06-25 · Public table; published model effort variant | 76.056 % benchmark maintainer ↗ | 78.873 % benchmark maintainer ↗ |
| LiveBench javascript · 2026-06-25 · Public table; published model effort variant | 68.182 % benchmark maintainer ↗ | 63.636 % benchmark maintainer ↗ |
| LiveBench AMPS_Hard · 2026-06-25 · Public table; published model effort variant | 98 % benchmark maintainer ↗ | 98 % benchmark maintainer ↗ |
| LiveBench olympiad · 2026-06-25 · Public table; published model effort variant | 91.286 % benchmark maintainer ↗ | 88.646 % benchmark maintainer ↗ |
| LiveBench theory_of_mind · 2026-06-25 · Public table; published model effort variant | 84.615 % benchmark maintainer ↗ | 73.077 % benchmark maintainer ↗ |
| LiveBench zebra_puzzle · 2026-06-25 · Public table; published model effort variant | 96.75 % benchmark maintainer ↗ | 93.5 % benchmark maintainer ↗ |
AS THE EVIDENCE CHANGED
Comparison history
Dated releases, price events, and benchmark publications. A price or test change can alter the factual comparison; this timeline does not pick a winner.
2026-08-13 · DeepSeek V4 Pro released ↗ · source
2026-07-09 · GPT-5.6 Luna released ↗ · source
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench code_completion (2026-06-25) at 86.957 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench code_generation (2026-06-25) at 78.873 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench javascript (2026-06-25) at 63.636 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench AMPS_Hard (2026-06-25) at 98 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench olympiad (2026-06-25) at 88.646 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench theory_of_mind (2026-06-25) at 73.077 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered LiveBench zebra_puzzle (2026-06-25) at 93.5 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered DeepSWE v1.1 maintainer pass@1 (1.1) at 67.1875 % · source ↗
2026-10-01 verified (test date not published) · deepseek-v4-pro entered Terminal-Bench 4.0 Vals mini-SWE-agent avg@3 fallbacks failed (4.0) at 14.141 % · source ↗
2026-10-01 verified (test date not published) · gpt-5.6-luna entered Terminal-Bench 4.0 Vals mini-SWE-agent avg@3 fallbacks failed (4.0) at 11.616 % · source ↗
2026-09-30 verified (test date not published) · deepseek-v4-pro entered LiveBench code_completion (2026-06-25) at 78.261 % · source ↗
2026-09-30 verified (test date not published) · deepseek-v4-pro entered LiveBench code_generation (2026-06-25) at 76.056 % · source ↗
An overall score appears only after enough current, comparable, independently sourced results exist. Pricing and individual benchmark results remain visible meanwhile.