ZMIME
Comparison · 2 models · Updated Oct 4, 2026

Grok 4.20 (Reasoning) vs Grok 4.3

Too close to call on our weighted score (Grok 4.20 (Reasoning) 68, Grok 4.3 67). The right pick depends on what you value most.

  1. xAI

    Grok 4.20 (Reasoning)

    Released Mar 9, 2026

    68/100
    • ECI152.0
    • Price$1.25 / $2.50
    • Context1M
    Full specs Remove Grok 4.20 (Reasoning)

  2. xAI

    Grok 4.3

    Released Apr 17, 2026

    67/100
    • ECI149.2
    • Price$1.25 / $2.50
    • Context1M
    Full specs Remove Grok 4.3

  3. Add a model

    Make it a three-way comparison.

01 — Verdict

Too close to call

It is close. Our weighted score puts them within 2 points (Grok 4.20 (Reasoning) 68/100, Grok 4.3 67/100), so choose by what matters most for your work: Grok 4.20 (Reasoning) for raw capability. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

  • CapabilityGrok 4.20 (Reasoning)Capabilities Index (ECI): Grok 4.20 (Reasoning) 152.0 · Grok 4.3 149.2
  • Lowest priceSame priceGrok 4.20 (Reasoning) $1.56 · Grok 4.3 $1.56 per 1M tokens (3:1 blend)
  • Longest contextAbout the sameGrok 4.20 (Reasoning) 1,000,000 · Grok 4.3 1,000,000 tokens
  • Widest inputsSame inputsGrok 4.20 (Reasoning): Text, Images, PDFs · Grok 4.3: Text, Images, PDFs
  • Self-hostingNo open weightsBoth are available only through APIs
How the score is built
MeasureWeightGrok 4.20 (Reasoning)Grok 4.3
CapabilityCapabilities Index (ECI)50%8177
Price25%4141
Inputs & features15%8080
Context window10%6060
Overall100%68/10067/100
02 — Side by side

Every spec in one table

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

Grok 4.20 (Reasoning) vs Grok 4.3 specifications side by side
SpecificationGrok 4.20 (Reasoning)xAIGrok 4.3xAI
Capability
Capabilities Index (ECI)152.0 (best)149.2
ECI rank#41 of 148 (best)#55 of 148
GPQA DiamondGraduate-level science questions89.3% (best)88.8%
FrontierMath Tiers 1–3Research-level mathematics44.9% (best)42.8%
OTIS Mock AIME 2024–2025Competition mathematics92.2%93.3% (best)
SimpleQA VerifiedShort factual questions30.2%33.2% (best)
Price per million tokens
Input$1.25$1.25
Output$2.50$2.50
Cached input$0.20$0.20
Blended (3:1)$1.56$1.56
Long-context rateOver 200K: $2.50 / $5.00Over 200K: $2.50 / $5.00
Price sourceOfficial xAI APIOfficial xAI API
Limits
Context window1,000,000 tokens1,000,000 tokens
Max output30,000 tokens30,000 tokens
Inputs and features
TextYesYes
ImagesYesYes
PDFsYesYes
AudioNoNo
VideoNoNo
ReasoningYesYeslow · medium · high
Tool callingYesYes
Structured outputYesYes
Availability
WeightsProprietaryProprietary
API model IDgrok-4.20-0309-reasoninggrok-4.3
API providers1127 (best)
ReleasedMar 9, 2026Apr 17, 2026
Knowledge cutoff——
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.

  • Grok 4.20 (Reasoning)$17.50
  • Grok 4.3$17.50
04 — Questions

Which should you choose?

Which is better: Grok 4.20 (Reasoning) or Grok 4.3?

It is close. Our weighted score puts them within 2 points (Grok 4.20 (Reasoning) 68/100, Grok 4.3 67/100), so choose by what matters most for your work: Grok 4.20 (Reasoning) for raw capability. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.

Which is cheaper, Grok 4.20 (Reasoning) or Grok 4.3?

Grok 4.20 (Reasoning) and Grok 4.3 cost the same: $1.25 input / $2.50 output per million tokens.

Which scores higher on benchmarks?

Grok 4.20 (Reasoning) scores higher on the Capabilities Index (ECI): Grok 4.20 (Reasoning) 152.0 (#41 of 148) and Grok 4.3 149.2 (#55 of 148). The confidence ranges of the top two overlap (149.3–154.4 vs 147.6–150.8), so treat the gap as small. On individual benchmarks: GPQA Diamond — Grok 4.20 (Reasoning) 89.3%, Grok 4.3 88.8%; FrontierMath Tiers 1–3 — Grok 4.20 (Reasoning) 44.9%, Grok 4.3 42.8%; OTIS Mock AIME 2024–2025 — Grok 4.3 93.3%, Grok 4.20 (Reasoning) 92.2%; SimpleQA Verified — Grok 4.3 33.2%, Grok 4.20 (Reasoning) 30.2%.

Which is better for coding?

There are no published SWE-bench Verified results for Grok 4.20 (Reasoning) and Grok 4.3 yet, so there is no like-for-like coding score. On overall capability, Grok 4.20 (Reasoning) leads, which tends to carry over to coding, but test on your own codebase. Both support tool calling for agent workflows.

Which has the bigger context window?

Grok 4.20 (Reasoning) and Grok 4.3 share the same 1,000,000-token context window. Maximum output per response: Grok 4.20 (Reasoning) up to 30,000, Grok 4.3 up to 30,000 tokens.

Which can read images, PDFs, audio or video?

Grok 4.20 (Reasoning) accepts text, images and PDFs; Grok 4.3 accepts text, images and PDFs. They handle the same number of input types.

Are any of these open source?

No. Grok 4.20 (Reasoning) and Grok 4.3 are proprietary and only available through APIs and apps.

Which is newer?

Grok 4.3 is the newest, released Apr 17, 2026. Grok 4.20 (Reasoning) came out Mar 9, 2026.

How do you decide the winner?

Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, 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 Oct 4, 2026.

05 — Keep comparing

Related comparisons