DeepSeekproprietary
Released: 2026-08-21Artificial AnalysisDeepSeek V4 Flash Vision
DeepSeek V4 Flash Vision from DeepSeek, tracked on the Artificial Analysis Intelligence Index.
Intelligence Index
—/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
159tok/sec
Output streaming throughputTime to First Answer
48s
Time to first answer token, including any thinkingCost to Run the Index
—USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of DeepSeek V4 Flash Vision (Max configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- —
- Cost to run the Index
- —
- Time per task
- —
- Omniscience index
- —
- Reasoning model
- Yes
- Cached-input price
- $0.014 / 1M
Reasoning & knowledge
- CritPt10.9%
- GPQA Diamond91.3%
- HLE-Diamond21.1%
- Humanity's Last Exam34.5%
Coding & terminal agents
- SciCode49.6%
- Terminal-Bench 2.174.2%
- Terminal-Bench 4.012.1%
Agentic work
- AA-Briefcase v1.1 (Elo)1418 Elo
- AutomationBench-AA47.5%
- EnterpriseOps-Gym-AA47.1%
- GDPval-AA (normalized Elo)1555 Elo
- τ³-Banking41.0%
Long context & documents
- AA-LCR (long-context reasoning)81.3%
- GDP.pdf12.0%
Multimodal & science
- MMMU-Pro74.8%
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Coding & agents · Terminal-Bench 4.012.1%
Long context · AA-LCR81.3%
Knowledge · Humanity's Last Exam34.5%
Sourced Benchmark Evaluations (13)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| GPQA | Nearing Saturation | 91.3% | 78 | 2026-11-15 | independent | Source ↗ |
| Humanity's Last Exam | Active | 34.5% | 93 | 2026-11-15 | independent | Source ↗ |
| MMMU-Pro | Active | 74.8% | 93 | 2026-11-15 | independent | Source ↗ |
| SciCode | Active | 49.7% | 93 | 2026-11-15 | independent | Source ↗ |
| τ³-Banking | Active | 41% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 2.1 | Active | 74.2% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 4.0 | Active | 12.1% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-LCR (Long Context Reasoning) | Active | 81.3% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt | Active | 10.9% | 93 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 21.1% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 12% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 47.5% | 93 | 2026-11-15 | independent | Source ↗ |
| EnterpriseOps-Gym-AA | Active | 47.1% | 93 | 2026-11-15 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="deepseek-deepseek-v4-flash-vision",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for DeepSeek V4 Flash Vision represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.