DeepSeek-V4-ProvsInkling
Across 6 shared benchmarks, Inkling leads overall: DeepSeek-V4-Pro wins 2, Inkling wins 4, with 0 ties and an average score difference of -4.73.
DeepSeek-V4-Pro
DeepSeek-AI · 2026-08-13 · Reasoning model
IN
Inkling
Thinking Machines Lab · 2026-07-15 · Multimodal model
DeepSeek-V4-Pro2 wins(33%)(67%)4 winsInkling
Benchmark scores
Grouped by capability, sorted by largest gap within each. 6 shared benchmarks.
Coding and Software Engineer
Inkling 2/2| Benchmark | DeepSeek-V4-Pro | Inkling | Diff |
|---|---|---|---|
| SWE-bench Verified | 73.6046 / 114Normal (With Tools) | 77.6026 / 114Thinking (With Tools) | -4 |
| SWE-Bench Pro - Public | 52.1040 / 57Normal (With Tools) | 54.3033 / 57Thinking (With Tools) | -2.20 |
AI Agent - Information Search
DeepSeek-V4-Pro 1/1| Benchmark | DeepSeek-V4-Pro | Inkling | Diff |
|---|---|---|---|
| BrowseComp | 83.4013 / 54极高强度思考（工具） | 77.1022 / 54Thinking (With Tools + Internet) | +6.30 |
AI Agent - Tool Usage
DeepSeek-V4-Pro 1/1| Benchmark | DeepSeek-V4-Pro | Inkling | Diff |
|---|---|---|---|
| Terminal-Bench 2.1 | 87.906 / 44极高强度思考（工具） | 63.8036 / 44Thinking (With Tools) | +24.10 |
General Evaluation
Inkling 1/1| Benchmark | DeepSeek-V4-Pro | Inkling | Diff |
|---|---|---|---|
| GPQA Diamond | 72.90147 / 226Normal (No Tools) | 87.2070 / 226Thinking (No Tools) | -14.30 |
General Knowledge
Inkling 1/1| Benchmark | DeepSeek-V4-Pro | Inkling | Diff |
|---|---|---|---|
| HLE | 7.70165 / 181Normal (No Tools) | 4641 / 181Thinking (With Tools) | -38.30 |
Specs
| Field | DeepSeek-V4-Pro | Inkling |
|---|---|---|
| Publisher | DeepSeek-AI | Thinking Machines Lab |
| Release date | 2026-08-13 | 2026-07-15 |
| Model type | Reasoning model | Multimodal model |
| Architecture | MoE | MoE |
| Parameters | 1.6T | 975B |
| Context length | 1M | 1M |
| Max output | 384K | Not available |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | DeepSeek-V4-Pro | Inkling |
|---|---|---|
| Text input | $0.435 / 1M tokens | $3.74 / 1M tokens |
| Text output | $0.87 / 1M tokens | $9.36 / 1M tokens |
| Cache read | $0.003625 / 1M tokens | $0.748 / 1M tokens |
Summary
- DeepSeek-V4-Proleads in:AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1)
- Inklingleads in:Coding and Software Engineer (2/2), General Evaluation (1/1), General Knowledge (1/1)
On average across the 6 shared benchmarks, Inkling scores 4.73 higher.
Largest single-benchmark gap: HLE — DeepSeek-V4-Pro 7.70 vs Inkling 46 (-38.30).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.