Across 4 comparable benchmarks, DeepSeek-V3-0324 has better reported results on 3 and worse on 1.
Moderate evidence4 of 4 shared results countedHow we compare
Price basis: Listed rates vs listed rates. The price section below shows the applicable periods.
| Fact | Llama 4 Maverick Instruct | DeepSeek-V3-0324 |
|---|---|---|
| Context window | 1M tokens | 128K tokens |
| Input / 1M | $0.240 | $1.14 |
| Output / 1M | $0.970 | $4.56 |
| Blended (4:1) | $0.386 | $1.82 |
| Status | Available | Available |
| Released | April 5, 2025 | March 24, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 1 shared benchmark grouped by capability on a 0–100 scale. How averages are built
DeepSeek-V3-0324 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
Llama 4 Maverick Instruct ahead on 1 of 1 counted
DeepSeek-V3-0324 ahead on 1 of 1 counted
No fixed-scale average — see the rows in Benchmark results
No shared comparable results for Abstract reasoning, Mathematics, Computer & tool use, Research & search, Long context, Multimodal, Cybersecurity.
Published scores, with matching modes where available. Gaps show Llama 4 Maverick Instruct relative to DeepSeek-V3-0324. Reading the results
| Benchmark | Llama 4 Maverick Instruct | DeepSeek-V3-0324 | Gap |
|---|---|---|---|
| LiveCodeBenchAlgorithmic Coding · Pass @K | 43.4Normal · no tools | 49.2Normal · no tools | 0.88×#91 vs #87 of 113 |
| GPQA DiamondScientific Reasoning · Accuracy | 69.8Normal · no tools | 68.4Normal · no tools | +1.4 pts#143 vs #147 of 200 |
| Creative WritingWriting | 858Normal · no tools | 1472Normal · no tools | 0.58×#109 vs #65 of 121 |
| MMLU-ProKnowledge Exams · Accuracy | 80.5Normal · no tools | 81.2Normal · no tools | −0.7 pts#59 vs #54 of 153 |
Algorithmic Coding · Pass @K
Scientific Reasoning · Accuracy
Writing
Knowledge Exams · Accuracy
A dash means no result is available.
| Benchmark | Llama 4 Maverick Instruct | DeepSeek-V3-0324 |
|---|---|---|
| AIME 2024 | — | 59.4 |
| AIME2025 | — | 47.7 |
| GSM8K | — | 96.3 |
| HLE | — | 5.2 |
| IMO-ProofBench | — | 4.3 |
| MATH-500 | — | 94.0 |
| MMLU | — | 86.5 |
| SciCode | — | 39.0 |
| Aider-Polyglot | — | 55.1 |
| IOI 2024 (Vals v1) | — | 1.7 |
| IOI 2025 (Vals v1) | — | 1.7 |
| SWE-bench Verified | — | 38.8 |
| Terminal Bench Hard | — | 15.2 |
| Terminal-Bench 1.0 | — | 13.3 |
| DROP | — | 89.7 |
| GDP.pdf | — | 1.6 |
| SAGE | 35.7 | — |
| τ²-Bench | — | 38.8 |
| τ³-Banking | — | 4.7 |
| ARC-AGI-1 | — | 9.0 |
| SimpleBench | — | 27.2 |
| GAIA | — | 29.4 |
| SimpleQA | — | 27.2 |
| MedCode | 36.5 | — |
| MedScribe | 54.2 | — |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
Llama 4 Maverick Instruct costs79%lessthan DeepSeek-V3-0324 per million tokens — 4.7× cheaper.Using listed rates vs listed rates at this token mix.
Llama 4 Maverick Instruct: Listed text rates; pricing tier not confirmed.
DeepSeek-V3-0324: Listed text rates; pricing tier not confirmed.
| Price / 1M tokens | Llama 4 Maverick Instruct | DeepSeek-V3-0324 | Ratio |
|---|---|---|---|
| Input | $0.24 | $1.14 | 0.21× |
| Output | $0.97 | $4.56 | 0.21× |
| Cached input | — | — | — |
| Cache write | — | — | — |
| Blended | $0.386 | $1.82 | 0.21× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Lower prices are highlighted. The cache hit rate applies to input tokens only; models without a published cached-input rate use their normal input price for those tokens.Effective dates: Llama 4 Maverick Instruct 2026-11-14; DeepSeek-V3-0324 2026-11-14.Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
| Attribute | Llama 4 Maverick Instruct | DeepSeek-V3-0324 |
|---|---|---|
| Availability | Available | Available |
| Provider | Facebook AI Research Lab | DeepSeek-AI |
| Model type | Multimodal model | Chat model |
| First release | April 5, 2025 | March 24, 2025 |
| Context window | 1M tokens | 128K tokens |
| Max output | 4K tokens | — |
| Total parameters | 400B | 671B |
| Active parameters | — | 37B |
| Architecture | Dense | Mixture of Experts (MoE) |
| Documented capabilities | Multilingual | Multilingual |
| Input modalities | Text, Image, Audio, Video | Text |
| Output modalities | Text | Text |
| Weights | Open weights | Open weights |
| Licence | Llama4 License | MIT License |
| Commercial use | Commercial use allowed | Commercial use allowed |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Llama 4 Maverick Instruct or DeepSeek-V3-0324? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.