ZMIME
Comparison · 3 models · Updated Nov 18, 2026

DeepSeek R1 0528 vs Grok 4.20 (Reasoning) vs Kimi K2.6

Grok 4.20 (Reasoning) comes out ahead, 77 to 74 and 57 on our weighted score, though DeepSeek R1 0528 is 40% cheaper per token.

Price in rating: Excluded · Include price

  1. DeepSeek

    DeepSeek R1 0528

    Released May 28, 2025

    57/100
    • ECI140.8
    • Price$0.512 / $2.19
    • Context164K
  2. Our pick

    xAI

    Grok 4.20 (Reasoning)

    Released Mar 9, 2026

    77/100
    • ECI152.3
    • Price$1.25 / $2.50
    • Context1M
  3. Moonshot AI

    Kimi K2.6

    Released Apr 20, 2026

    74/100
    • ECI151.6
    • Price$0.95 / $4.00
    • Context262K
01 — Verdict

Grok 4.20 (Reasoning) is our pick

Grok 4.20 (Reasoning) is the better all-round choice, scoring 77/100 against Kimi K2.6 (74) and DeepSeek R1 0528 (57). It leads on context window. The score weighs capability 67%, inputs & features 20%, context window 13%.

  • CapabilityGrok 4.20 (Reasoning)Capabilities Index (ECI): Grok 4.20 (Reasoning) 152.3 · Kimi K2.6 151.6 · DeepSeek R1 0528 140.8
  • Lowest priceDeepSeek R1 0528DeepSeek R1 0528 $0.932 · Grok 4.20 (Reasoning) $1.56 · Kimi K2.6 $1.71 per 1M tokens (3:1 blend)
  • Longest contextGrok 4.20 (Reasoning)Grok 4.20 (Reasoning) 1,000,000 · Kimi K2.6 262,144 · DeepSeek R1 0528 163,840 tokens
  • Widest inputsGrok 4.20 (Reasoning) and Kimi K2.6DeepSeek R1 0528: Text · Grok 4.20 (Reasoning): Text, Images, PDFs · Kimi K2.6: Text, Images, Video
  • Self-hostingDeepSeek R1 0528 and Kimi K2.6Publishes downloadable weights (MIT)
How the score is built
MeasureWeightDeepSeek R1 0528Grok 4.20 (Reasoning)Kimi K2.6
CapabilityCapabilities Index (ECI)67%697979
Inputs & features20%358080
Context window13%366443
Overall100%57/10077/10074/100
02 — Side by side

Every spec in one table

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

DeepSeek R1 0528 vs Grok 4.20 (Reasoning) vs Kimi K2.6 specifications side by side
SpecificationDeepSeek R1 0528DeepSeekGrok 4.20 (Reasoning)xAIKimi K2.6Moonshot AI
Capability
Capabilities Index (ECI)140.8152.3 (best)151.6
ECI rank#113 of 158#59 of 158 (best)#61 of 158
GPQA DiamondGraduate-level science questions76.3%89.3%90.8% (best)
FrontierMath Tiers 1–3Research-level mathematics—44.9%57.2% (best)
OTIS Mock AIME 2024–2025Competition mathematics66.4%92.2%96.1% (best)
SWE-bench VerifiedFixing real GitHub issues——76.7%
SimpleQA VerifiedShort factual questions—30.2%34.9% (best)
Price per million tokens
Input$0.512 (best)$1.25$0.95
Output$2.19 (best)$2.50$4.00
Cached input—$0.20$0.16 (best)
Blended (3:1)$0.932 (best)$1.56$1.71
Long-context rateSame rateOver 200K: $2.50 / $5.00Same rate
Price sourceMedian of 16 providersOfficial xAI APIOfficial Moonshot AI API
Limits
Context window163,840 tokens1,000,000 tokens (best)262,144 tokens
Max output32,768 tokens30,000 tokens262,144 tokens (best)
Inputs and features
TextYesYesYes
ImagesNoYesYes
PDFsNoYesNo
AudioNoNoNo
VideoNoNoYes
ReasoningYesYesYes
Tool callingYesYesYes
Structured outputNoYesYes
Availability
WeightsOpenMITProprietaryOpen
API model ID—grok-4.20-0309-reasoningkimi-k2.6
API providers161464 (best)
ReleasedMay 28, 2025Mar 9, 2026Apr 20, 2026
Knowledge cutoff——Jan 2025
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 0528$9.50
  • Grok 4.20 (Reasoning)$17.50
  • Kimi K2.6$17.50
04 — Questions

Which should you choose?

Which is better: DeepSeek R1 0528, Grok 4.20 (Reasoning) or Kimi K2.6?

Grok 4.20 (Reasoning) is the better all-round choice, scoring 77/100 against Kimi K2.6 (74) and DeepSeek R1 0528 (57). It leads on context window. The score weighs capability 67%, inputs & features 20%, context window 13%.

Which is cheaper, DeepSeek R1 0528, Grok 4.20 (Reasoning) or Kimi K2.6?

DeepSeek R1 0528 is cheaper at $0.512 input / $2.19 output per million tokens (median across 16 API providers). Grok 4.20 (Reasoning) costs $1.25 input / $2.50 output per million tokens (official xAI API price); Kimi K2.6 costs $0.95 input / $4.00 output per million tokens (official Moonshot AI API price). At a typical mix of three input tokens to one output token, that is $0.932 per million tokens for DeepSeek R1 0528 versus $1.56 for Grok 4.20 (Reasoning) (1.7× as much) and $1.71 for Kimi K2.6 (1.8× as much).

Which scores higher on benchmarks?

Grok 4.20 (Reasoning) scores higher on the Capabilities Index (ECI): Grok 4.20 (Reasoning) 152.3 (#59 of 158), Kimi K2.6 151.6 (#61 of 158) and DeepSeek R1 0528 140.8 (#113 of 158). The confidence ranges of the top two overlap (150.6–154.1 vs 149.9–153.4), so treat the gap as small. On individual benchmarks: GPQA Diamond — Kimi K2.6 90.8%, Grok 4.20 (Reasoning) 89.3%, DeepSeek R1 0528 76.3%; OTIS Mock AIME 2024–2025 — Kimi K2.6 96.1%, Grok 4.20 (Reasoning) 92.2%, DeepSeek R1 0528 66.4%.

Which is better for coding?

There are no published SWE-bench Verified results for DeepSeek R1 0528 and Grok 4.20 (Reasoning) 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. All three support tool calling for agent workflows.

Which has the bigger context window?

Grok 4.20 (Reasoning) has the largest context window at 1,000,000 tokens, against 262,144 for Kimi K2.6 and 163,840 for DeepSeek R1 0528. Maximum output per response: DeepSeek R1 0528 up to 32,768, Grok 4.20 (Reasoning) up to 30,000, Kimi K2.6 up to 262,144 tokens.

Which can read images, PDFs, audio or video?

DeepSeek R1 0528 accepts text; Grok 4.20 (Reasoning) accepts text, images and PDFs; Kimi K2.6 accepts text, images and video. Grok 4.20 (Reasoning) handles the widest range of inputs.

Are any of these open source?

DeepSeek R1 0528 and Kimi K2.6 publishes its weights (MIT) and can be self-hosted; Grok 4.20 (Reasoning) is proprietary.

Which is newer?

Kimi K2.6 is the newest, released Apr 20, 2026. Grok 4.20 (Reasoning) came out Mar 9, 2026; DeepSeek R1 0528 came out May 28, 2025. Knowledge cutoff: Kimi K2.6 Jan 2025.

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.