DeepSeek V3-0324
DeepSeek V3-0324, the March 2025 update of DeepSeek V3, served under DeepSeek's original deepseek-chat name. Quantized at FP8.
- Native PDF input
- Tool Calling
- Structured Output
Added Feb 27, 2025
Model weightsPricing
Auto routing · per 1M tokens- Input
- $0.10
- Output
- $0.425
- Cache read
- $0.05
Specifications
- Context window
- 128K
- Max output
- 8.2K
- Parameters
- 671B / 37B
- Total / active
Benchmarks
Benchmarks
Sourced from Vectara.
Hallucination Rate
6.1%
Factual Consistency
93.9%
Answer Rate
97.5%
Avg Summary Length
Average generated summary length
81.7
Last updated 2026-10-22 · Matched as deepseek-ai/DeepSeek-V3
Vectara LeaderboardLeaderboard results
These leaderboards tested DeepSeek Chat 0324, another version of this model.
| Benchmark | Score | Rank on NanoGPT |
|---|---|---|
Epoch AI Benchmarking Hub Epoch AI data, CC BY 4.0 · Updated Nov 13, 2026 | ||
GPQA Diamond | 67.6% | #93 of 119 |
OTIS Mock AIME 2024–2025 | 37.8% | #106 of 121 |
Creative Writing v3 | 1472 | #61 of 96 |
Longform Writing | 60.1 | #55 of 97 |
EQ-Bench 3 | 1124 | #32 of 49 |
UGI Leaderboard UGI Leaderboard data, Apache-2.0 · Updated Oct 16, 2026 | ||
UGI Score | 55.0 | #23 of 149 |
UGI Willingness | 5.8 | #30 of 149 |
UGI Writing | 49.1 | #61 of 149 |
Filtered and ranked to models on NanoGPT.
Providers
Auto routing is available for this model. Explicit provider selection is not available.
 Recommended | $0.20 | $0.77 | $0.135 | N/A | N/A |
Batch rows for this model use normal routing and cost the same as standard API requests. Set provider in a row to choose a provider.