DeepSeek V3 API Benchmarks, Pricing & Provider Data
Compare DeepSeek V3 with another model
Choose a model to open its comparison page.
LLM
DeepSeek V3 benchmark, API pricing, and provider data cover 511 API providers, with prices starting at $0.0037/M. DeepSeek V3 free API options are available from 7 providers. The page also shows measured API speed and first-token latency.
DeepSeek V3 is DeepSeek flagship MoE language model with 671B total parameters, delivering strong performance in reasoning, coding, and multilingual tasks at competitive inference cost.
Quality
#87of 100
44.0
LMSpeed score
Speed
49char/s
3.39 s
Cost
#70of 180
$0.0037/ 1M · 8:1 in:out
$0.0006 in · $0.0031 out
Category Performance
Observed-capability estimates with uncertainty reported separately from independent benchmark families.
- Coverage
- 6 / 8
- Methodology
- V3.0
#1Math49.5Estimated3/4 Measured dimensions
80% interval: 40.2–58.7
foundational math54.7
math_500 · math_500
competition math53.6
aime · aime
proof frontier40.1
frontiermath · frontier_math_v2_tiers13
applied tool mathPrior only
#2Reasoning44.4RatedGlobal rank #563/4 Measured dimensions
80% interval: 35.7–53.0
abstract logicPrior only
scientific causal43.7
critpt · critpt / gpqa · gpqa / hle · hle
multistep constraints53.2
mmlu_pro · mmlu_pro
evidence verification36.2
lcr · lcr
#3Knowledge41.1Provisional1/4 Measured dimensions
80% interval: 25.1–57.1
broad knowledge41.1
aa_omniscience · aa_omniscience_index
professional knowledgePrior only
factualityPrior only
retrieval open bookPrior only
#4Instruction following39.7Estimated2/4 Measured dimensions
80% interval: 27.8–51.6
constraint following37.7
ifeval · ifeval
structured outputPrior only
novel instruction generalization41.8
ifbench · aa_if_bench
long multiturn instructionPrior only
#5Coding37.4Estimated2/4 Measured dimensions
80% interval: 26.2–48.5
code generation46.9
livecodebench · livecodebench / scicode · scicode
repository engineeringPrior only
debugging testing27.9
swe_verified · swe_verified
tooling qualityPrior only
#6Agents34.5Estimated2/4 Measured dimensions
80% interval: 22.7–46.3
planningPrior only
tool use38.5
tau · tau2_bench
environment execution30.5
gdpval_aa · benchlm_agentic_gdpval_aa
recovery reliabilityPrior only
No data:MultilingualMultimodal
Specifications
Rankings
Excels at
It's decent at
Falls behind in
Detailed scores
Updated: Aug 31, 2026Overall
undefined metric} other undefined metrics}}
Overall score44.0#87 / 100
Overall
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Pricing
undefined metric} other undefined metrics}}
Input price$0.360/M#70 / 180Output price$0.890/M#42 / 180
Pricing
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Agents
V3.0undefined metric} other undefined metrics}} · Estimated
Score34.580% interval22.7–46.32/4 Measured dimensionsAA Agentic Index1.6#61 / 65Τ²-bench results22.8#76 / 82
Agents
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Planning & decompositionPrior only
Tool use38.5
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
tau · tau2_bench · z -1.65 · q 1.00
Environment & long-horizon execution30.5
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
gdpval_aa · benchlm_agentic_gdpval_aa · z -2.77 · q 1.00
Recovery & completion reliabilityPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Coding
V3.0undefined metric} other undefined metrics}} · Estimated
Score37.480% interval26.2–48.52/4 Measured dimensionsLiveCodeBench35.9%#78 / 115SciCode35.4%#127 / 205
Coding
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Code generation46.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
livecodebench · livecodebench · z -0.21 · q 1.00
scicode · scicode · z -0.20 · q 1.00
Repository engineeringPrior only
Debugging & testing27.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
swe_verified · swe_verified · z -3.00 · q 1.00
Tool-assisted development & qualityPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
LiveCodeBenchV3 evidence
LMSpeed rank#78
Score35.9%
Sourceartificialanalysis.ai
UpdatedAug 31, 2026
Confidence4
SciCodeV3 evidence
LMSpeed rank#127
Score35.4%
Sourceartificialanalysis.ai
UpdatedAug 31, 2026
Confidence4
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Score44.4#5680% interval35.7–53.03/4 Measured dimensionsMMLU-Pro75.2%#85 / 129GPQA55.7%#177 / 212
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Abstract logicPrior only
Scientific & causal reasoning43.7
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
critpt · critpt · z -0.72 · q 1.00
gpqa · gpqa · z -0.55 · q 1.00
hle · hle · z -0.77 · q 1.00
Multi-step constraints53.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mmlu_pro · mmlu_pro · z 0.22 · q 1.00
Evidence integration & verification36.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
lcr · lcr · z -2.19 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
MMLU-ProV3 evidence
LMSpeed rank#85
Score75.2%
Sourceartificialanalysis.ai
UpdatedAug 31, 2026
Confidence4
Knowledge
V3.0undefined metric} other undefined metrics}} · Provisional
Score41.180% interval25.1–57.11/4 Measured dimensionsArtificial Analysis Intelligence Index14.2#96 / 111AA-GPQA Diamond55.7#101 / 108
Knowledge
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Broad knowledge41.1
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
aa_omniscience · aa_omniscience_index · z -1.27 · q 1.00
Professional knowledgePrior only
FactualityPrior only
Retrieval & open-book usePrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Artificial Analysis Intelligence IndexV3 evidence
LMSpeed rank#96
Score14.2
Sourcebenchlm.ai
UpdatedAug 20, 2026
Confidence1
AA-Omniscience Hallucination Rate
LMSpeed rank#19
Score90.0
Sourcebenchlm.ai
UpdatedAug 20, 2026
Confidence1
Math
V3.0undefined metric} other undefined metrics}} · Estimated
Score49.580% interval40.2–58.73/4 Measured dimensionsAIME25.3%#44 / 68MATH-50088.7%#37 / 73
Math
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Foundational math54.7
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
math_500 · math_500 · z 0.66 · q 1.00
Competition math53.6
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
aime · aime · z 0.46 · q 1.00
Advanced proofs40.1
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
frontiermath · frontier_math_v2_tiers13 · z -1.67 · q 1.00
Applied & tool-assisted mathPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
MATH-500V3 evidence
LMSpeed rank#37
Score88.7%
Sourceartificialanalysis.ai
UpdatedAug 31, 2026
Confidence4
FrontierMath v2 (Tiers 1-3)V3 evidence
LMSpeed rank#44
Score1.7
Sourcebenchlm.ai
UpdatedAug 20, 2026
Confidence1
Multilingual
V3.0undefined metric} other undefined metrics}} · No data
No data80% interval30.8–69.20/4 Measured dimensions
Multilingual
V3.0undefined metric} other undefined metrics}} · No data
Dimensions and evidence
Cross-language understandingPrior only
Multilingual generationPrior only
Reasoning transferPrior only
Low-resource robustnessPrior only
Multimodal
V3.0undefined metric} other undefined metrics}} · No data
80% interval30.8–69.20/4 Measured dimensionsDesign Arena Website1131.0#63 / 75
Multimodal
V3.0undefined metric} other undefined metrics}} · No data
Dimensions and evidence
Perception & OCRPrior only
Document & spatial understandingPrior only
Visual reasoningPrior only
Video & grounded actionPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Instruction following
V3.0undefined metric} other undefined metrics}} · Estimated
Score39.780% interval27.8–51.62/4 Measured dimensionsIFEval86.1#15 / 16AA-IFBench34.8#79 / 84
Instruction following
V3.0undefined metric} other undefined metrics}} · Estimated
Dimensions and evidence
Constraint following37.7
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
ifeval · ifeval · z -1.84 · q 1.00
Structured outputPrior only
Novel-instruction generalization41.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
ifbench · aa_if_bench · z -1.37 · q 1.00
Multi-turn & long instructionsPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Pricing Comparison
Compare DeepSeek V3 API pricing across 504 providers. Prices range from $0.0037/M to $95.45/M. 6345ywz API offers the lowest rate at $0.0037/M. 7 providers offer free API credits or a free tier.
| Provider | Health | Model Variant | Group | Input ($/M) | Output ($/M) | Speed (t/s) | First token | Audit |
|---|---|---|---|---|---|---|---|---|
L1 100% | deepseek-v3 | default | -62%$0.137/M | -38%$0.548/M | — | — | — | |
L1 100% | deepseek-v3-0324 | default | -62%$0.137/M | -38%$0.548/M | — | — | — | |
L1 100% | deepseek-v3-250324 | default | -62%$0.137/M | -38%$0.548/M | — | — | — | |
L1 100% | DeepSeek-V3 | default | $0.025/request | - | — | — | — | |
L1 100% | deepseek-v3-0324 | default | -24%$0.274/M | $1.10/M | — | — | — | |
L1 100% | deepseek-v3 | default | -24%$0.274/M | $1.10/M | — | — | — | |
L1 100% | deepseek-v3-fast | default | $10.27/M | $20.55/M | — | — | — | |
L1 100% | siliconflow/deepseek-v3-0324 | default | $10.27/M | $10.27/M | — | — | — | |
L1 100% | deepseek-v3-0324 | default | -45%$0.197/M | -11%$0.789/M | — | — | — | |
L1 100% | deepseek-v3 | default | -40%$0.215/M | -3%$0.859/M | — | — | — | |
L1 100% | deepseek-v3-0324 | default | -24%$0.274/M | $1.10/M | — | — | — | |
L1 100% | deepseek-v3 | default | -24%$0.274/M | $1.10/M | — | — | — | |
L1 100% | deepseek-v3-250324 | default | -24%$0.274/M | $1.10/M | — | — | — | |
L1 100% | deepseek-ai/DeepSeek-V3 | default | $0.600/M | $2.40/M | — | — | — | |
兔子API Free | L1 100% | deepseek-v3-all | default | Free | Free | — | — | — |
L1 100% | deepseek-v3 | default | Free | Free | — | — | — | |
L1 100% | deepseek-v3-0324 | default | Free | Free | — | — | — | |
L1 100% | deepseek-v3-250324 | default | Free | Free | — | — | — | |
L1 100% | deepseek-v3-0324 | default | $0.010/request | - | — | — | — | |
L1 100% | deepseek-v3-0324 | default | -77%$0.082/M | -63%$0.329/M | — | — | — |
Showing 20 model IDs of 127.
Alternatives & Similar Models
DeepSeek R1
deepseek-r1
DeepSeek R1 is a reasoning-focused language model in the DeepSeek series, designed for complex reasoning, problem-solving, and analytical tasks.
156 shared providers
DeepSeek V3.2
deepseek-v3-2
DeepSeek V3.2 is an upgraded V3-series MoE model with stronger reasoning, coding, and math performance, widely available through OpenAI-compatible API relays.
147 shared providers
Kimi K2.5
kimi-k2-5
Moonshot Kimi K2.5 is an open-weight multimodal agent model with native vision and text input, strong coding performance, and a 256K context window.
134 shared providers
GLM-5
glm-5
Zhipu GLM-5 is Zhipu flagship GLM series model with enhanced reasoning, agent capabilities, and strong performance on Chinese enterprise and coding scenarios.
132 shared providers
DeepSeek V3.1
deepseek-v3-1
DeepSeek V3.1 is an open-weights style frontier model from DeepSeek with strong math, coding, and Chinese-English bilingual reasoning.
129 shared providers
DeepSeek V4 Flash
deepseek-v4-flash
DeepSeek V4 Flash is a fast, cost-efficient language model in the DeepSeek V4 family, optimized for low-latency chat, coding assistance, and high-throughput API workloads while retaining strong reasoning quality.
128 shared providers
Frequently Asked Questions
- What benchmark data does DeepSeek V3 include?
- LMSpeed shows DeepSeek V3 benchmark context, API price, output speed, first-token latency, and provider data across 511 providers when those signals are available.
- What is the DeepSeek V3 API price?
- DeepSeek V3 has pricing from 511 providers, ranging from $0.0037/M to $95.45/M. 6345ywz API has the lowest listed price.
- What does the DeepSeek V3 API pricing table include?
- The DeepSeek V3 API pricing table compares 511 providers by input price, output price, free tier status, speed, first-token latency, and recent health data when available.
- Which provider has the cheapest DeepSeek V3 API pricing?
- 6345ywz API currently has the lowest listed DeepSeek V3 price at $0.0037/M across 511 providers.
- Can I compare DeepSeek V3 API price and speed together?
- Yes. LMSpeed shows DeepSeek V3 API price, output speed, first-token latency, and provider health on the same page so you can compare cost and performance together.
- Is DeepSeek V3 API free?
- Yes, DeepSeek V3 free API options are available through 7 providerundefined other undefined} on LMSpeed, including 兔子API, 兔子API, 兔子API, 兔子API, 兔子API. These providers offer free API credits or a free tier with no per-token charges.
- Where can I get DeepSeek V3 free API access?
- LMSpeed currently lists 7 free API providerundefined other undefined} for DeepSeek V3: 兔子API, 兔子API, 兔子API, 兔子API, 兔子API. Check each provider row before using it because free tier limits can change.