ZMIME
Comparison · 3 models · Updated Nov 18, 2026

DeepSeek-R1 vs DeepSeek-V3 vs Llama 4 Maverick 17B Instruct

Llama 4 Maverick 17B Instruct comes out ahead, 59 to 56 and 49 on our weighted score, and it is the cheaper option too.

Price in rating: Excluded · Include price

  1. DeepSeek

    DeepSeek-R1

    Released Jan 20, 2025

    56/100
    • ECI138.6
    • Price$1.02 / $5.14
    • Context128K
  2. DeepSeek

    DeepSeek-V3

    Released Dec 26, 2024

    49/100
    • ECI131.7
    • Price$0.32 / $1.10
    • Context131K
  3. Our pick

    Meta

    Llama 4 Maverick 17B Instruct

    Released Apr 5, 2025

    59/100
    • ECI132.2
    • Price$0.27 / $0.86
    • Context1M
01 — Verdict

Llama 4 Maverick 17B Instruct is our pick

Llama 4 Maverick 17B Instruct is the better all-round choice, scoring 59/100 against DeepSeek-R1 (56) and DeepSeek-V3 (49). It leads on inputs & features and context window. DeepSeek-R1 wins on capability. The score weighs capability 67%, inputs & features 20%, context window 13%.

  • CapabilityDeepSeek-R1Capabilities Index (ECI): DeepSeek-R1 138.6 · Llama 4 Maverick 17B Instruct 132.2 · DeepSeek-V3 131.7
  • Lowest priceLlama 4 Maverick 17B InstructLlama 4 Maverick 17B Instruct $0.417 · DeepSeek-V3 $0.515 · DeepSeek-R1 $2.05 per 1M tokens (3:1 blend)
  • Longest contextLlama 4 Maverick 17B InstructLlama 4 Maverick 17B Instruct 1,000,000 · DeepSeek-V3 131,072 · DeepSeek-R1 128,000 tokens
  • Widest inputsLlama 4 Maverick 17B InstructDeepSeek-R1: Text · DeepSeek-V3: Text · Llama 4 Maverick 17B Instruct: Text, Images
  • Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
How the score is built
MeasureWeightDeepSeek-R1DeepSeek-V3Llama 4 Maverick 17B Instruct
CapabilityCapabilities Index (ECI)67%676061
Inputs & features20%352550
Context window13%323264
Overall100%56/10049/10059/100
02 — Side by side

Every spec in one table

Highlighted cells lead their row. Dashes mean the data is not published.

DeepSeek-R1 vs DeepSeek-V3 vs Llama 4 Maverick 17B Instruct specifications side by side
SpecificationDeepSeek-R1DeepSeekDeepSeek-V3DeepSeekLlama 4 Maverick 17B InstructMeta
Capability
Capabilities Index (ECI)138.6 (best)131.7132.2
ECI rank#118 of 158 (best)#136 of 158#134 of 158
GPQA DiamondGraduate-level science questions71.7% (best)56.5%67.0%
OTIS Mock AIME 2024–2025Competition mathematics53.3% (best)15.8%20.6%
Price per million tokens
Input$1.02$0.32$0.27 (best)
Output$5.14$1.10$0.86 (best)
Cached input———
Blended (3:1)$2.05$0.515$0.417 (best)
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 12 providersMedian of 9 providersMedian of 16 providers
Limits
Context window128,000 tokens131,072 tokens1,000,000 tokens (best)
Max output32,768 tokens (best)8,192 tokens16,384 tokens
Inputs and features
TextYesYesYes
ImagesNoNoYes
PDFsNoNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningYesNoNo
Tool callingYesYesYes
Structured outputNoNoNo
Availability
WeightsOpenOpenDeepSeek Model LicenseOpen
API model ID———
API providers13916 (best)
ReleasedJan 20, 2025Dec 26, 2024Apr 5, 2025
Knowledge cutoffJul 2024—Aug 2024
03 — Cost

What would a month cost?

Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.

  • DeepSeek-R1$20.53
  • DeepSeek-V3$5.40
  • Llama 4 Maverick 17B Instruct$4.42
04 — Questions

Which should you choose?

Which is better: DeepSeek-R1, DeepSeek-V3 or Llama 4 Maverick 17B Instruct?

Llama 4 Maverick 17B Instruct is the better all-round choice, scoring 59/100 against DeepSeek-R1 (56) and DeepSeek-V3 (49). It leads on inputs & features and context window. DeepSeek-R1 wins on capability. The score weighs capability 67%, inputs & features 20%, context window 13%.

Which is cheaper, DeepSeek-R1, DeepSeek-V3 or Llama 4 Maverick 17B Instruct?

Llama 4 Maverick 17B Instruct is cheaper at $0.27 input / $0.86 output per million tokens (median across 16 API providers). DeepSeek-V3 costs $0.32 input / $1.10 output per million tokens (median across 9 API providers); DeepSeek-R1 costs $1.02 input / $5.14 output per million tokens (median across 12 API providers). At a typical mix of three input tokens to one output token, that is $0.417 per million tokens for Llama 4 Maverick 17B Instruct versus $0.515 for DeepSeek-V3 (1.2× as much) and $2.05 for DeepSeek-R1 (4.9× as much).

Which scores higher on benchmarks?

DeepSeek-R1 scores higher on the Capabilities Index (ECI): DeepSeek-R1 138.6 (#118 of 158), Llama 4 Maverick 17B Instruct 132.2 (#134 of 158) and DeepSeek-V3 131.7 (#136 of 158). Their confidence ranges do not overlap (136.7–140.4 vs 130.0–134.3), so the gap is a real one. On individual benchmarks: GPQA Diamond — DeepSeek-R1 71.7%, Llama 4 Maverick 17B Instruct 67.0%, DeepSeek-V3 56.5%; OTIS Mock AIME 2024–2025 — DeepSeek-R1 53.3%, Llama 4 Maverick 17B Instruct 20.6%, DeepSeek-V3 15.8%.

Which is better for coding?

There are no published SWE-bench Verified results for DeepSeek-R1, DeepSeek-V3 and Llama 4 Maverick 17B Instruct yet, so there is no like-for-like coding score. On overall capability, DeepSeek-R1 leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.

Which has the bigger context window?

Llama 4 Maverick 17B Instruct has the largest context window at 1,000,000 tokens, against 131,072 for DeepSeek-V3 and 128,000 for DeepSeek-R1. Maximum output per response: DeepSeek-R1 up to 32,768, DeepSeek-V3 up to 8,192, Llama 4 Maverick 17B Instruct up to 16,384 tokens.

Which can read images, PDFs, audio or video?

DeepSeek-R1 accepts text; DeepSeek-V3 accepts text; Llama 4 Maverick 17B Instruct accepts text and images. Llama 4 Maverick 17B Instruct handles the widest range of inputs.

Are any of these open source?

Yes, all three publish their weights (DeepSeek Model License), so you can self-host them.

Which is newer?

Llama 4 Maverick 17B Instruct is the newest, released Apr 5, 2025. DeepSeek-R1 came out Jan 20, 2025; DeepSeek-V3 came out Dec 26, 2024. Knowledge cutoff: DeepSeek-R1 Jul 2024, Llama 4 Maverick 17B Instruct Aug 2024.

How do you decide the winner?

This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 18, 2026.