Model comparison · accepted catalogue facts 2026-11-22

Grok 4.3 vs Grok 4.6

Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.

Both models belong to the xAI Grok family. Inclusion does not recommend either model. This is a direct pair comparison; global frontier membership is not its verdict.

Grok 4.3

LMArena general · Elo · higher is better

Scale starts at 1040 Elo

1396.8

Balanced effective USD / 1M tokens · lower is better

$1.56/M

Grok 4.6

LMArena general · Elo · higher is better

Scale starts at 1040 Elo

1430.0

Balanced effective USD / 1M tokens · lower is better

$3.00/M

In-Depth Analysis

Grok 4.3 vs Grok 4.6: Measured Technical Breakdown

Comprehensive comparison across measured human-preference evaluation, effective production token costs, architectural limits, and migration trade-offs.

Measured Capability (LMArena)
Grok 4.6 leads by 33.2 Elo

Grok 4.3: 1396.8 ±3.63 · Grok 4.6: 1430 ±4.67 Statistically significant preference separation

Cost Comparison Across Canonical Production Workloads

Rates represent effective cost per 1M tokens under standard token blends. Cached prompt reads are factored according to workload context reuse.

Workload & RatioGrok 4.3Grok 4.6Cost DifferenceAdvantage
Balanced 3 tokens in per 1 out$1.56 /1M tokens$3.00 /1M tokens$1.44 (48% diff)Grok 4.3 costs less
Summarise Huge input, tiny output$1.01 /1M tokens$1.77 /1M tokens$0.759 (43% diff)Grok 4.3 costs less
Chat Long cached system prompt$1.44 /1M tokens$3.15 /1M tokens$1.72 (54% diff)Grok 4.3 costs less
Code gen Output dominates$1.92 /1M tokens$4.28 /1M tokens$2.36 (55% diff)Grok 4.3 costs less
Agentic Context replayed each turn$0.813 /1M tokens$1.71 /1M tokens$0.895 (52% diff)Grok 4.3 costs less

Architectural & Operational Specifications

SpecificationGrok 4.3Grok 4.6Operational Note
Context window1,000,000 tokens500,000 tokensGrok 4.3 has a 2.0× larger context window
Max output generation900,000 tokens450,000 tokensGrok 4.3 allows longer single-turn completions
Base input price$1.25 / 1M tokens$2.00 / 1M tokensGrok 4.3 input rate is cheaper
Base output price$2.50 / 1M tokens$6.00 / 1M tokensGrok 4.3 output rate is cheaper
Cached input price$0.200 / 1M tokens$0.500 / 1M tokensGrok 4.3 offers lower cached prompt cost
Reasoning & thinking tokensSupported; standard ratesSupported; standard rates—
Input modalitiestext, image, filetext, image, file—
Tool calling / function useSupportedSupported—
Open weights & licenceProprietary (proprietary)Proprietary (proprietary)—
Provider & reseller marketxAI · 4 provider offersxAI · 4 provider offers—

When to Choose Grok 4.3

  • Larger context window (1,000,000 tokens) for long-document or repo-level context.
  • Higher maximum completion limit (900,000 tokens).
  • Direct integration with xAI tooling and API features.

When to Choose Grok 4.6

  • Higher measured general capability (1430 Elo on human evaluation).
  • Direct integration with xAI tooling and API features.
Risks when moving from Grok 4.3 → Grok 4.6:
  • Context window: 1,000,000 → 500,000 tokens.
  • Output limit: 900,000 → 450,000 tokens.

Compare · model decision

Choose two or three models

Inspect the differences, estimate your workload and keep the evidence. All standard catalogue models are available, including unrated ones.

Catalogue facts
2026-11-22

2 / 3 selected · 0 requirements
Open public comparison link
Public workload and requirements

These settings are included in the share link. The fixed-request scenario and private evaluation data below are excluded.

A ↔ B

Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.

Estimated blend: 75% input, 25% output; 0% cached input. Base context tier. Prompt length selects the billing tier. An unlisted cache-read rate uses that tier’s full input rate in this blend; missing required input or output rates prevent an estimate. The fixed-request estimate below validates actual context and output limits and requires a published cache-read rate when cache hits are assumed.

What changes between these models

Quality first, then estimated price. LMArena scores are human-preference measurements, not a benchmark of your application. A missing score is unrated.

  • Grok 4.3 · Meets the stated model requirements
  • Grok 4.6 · Meets the stated model requirements
Material differences and unknown evidence. Deltas compare each model with model A. Unknown facts remain visible.
Measured factA Grok 4.3B Grok 4.6
LMArena general score (Elo)1,396.81,430Δ vs A: +33.2 Elo
Score interval half-width (± Elo)3.634.67Δ vs A: +1.04 Elo
Benchmark reasoning effortdefaulthigh
LMArena document score (Elo)Unknown1,438.1
Estimated effective USD / 1M tokens$1.5625$3Δ vs A: +1.4375 USD/M
Base input USD / 1M tokens$1.25$2Δ vs A: +0.75 USD/M
Base output USD / 1M tokens$2.5$6Δ vs A: +3.5 USD/M
Base cached-input USD / 1M tokens$0.2$0.5Δ vs A: +0.3 USD/M
Context window (tokens)1,000,000500,000Δ vs A: -500,000 tokens
Output limit (tokens)900,000450,000Δ vs A: -450,000 tokens
Recorded retirement dateUnknownUnknown
9 shared measured facts
Shared measured facts. Deltas compare each model with model A. Unknown facts remain visible.
Measured factA Grok 4.3B Grok 4.6
Input modalitiestext, image, filetext, image, file
Tool useSupportedSupportedΔ vs A: 0
ReasoningSupportedSupportedΔ vs A: 0
Open weightsNot supportedNot supportedΔ vs A: 0
Recorded licenceproprietaryproprietary
Model providerxAIxAI
Distinct sellers in the catalogue11Δ vs A: 0
Serving precision differs across offersNot supportedNot supportedΔ vs A: 0
Retirement evidenceNone announced; not a guaranteeNone announced; not a guarantee

Provider counts and precision flags describe catalogue offers. They do not establish the precision, availability, region, tool behaviour or limits of your chosen endpoint. Open each model’s provider table before switching.

Complete price ladders and billing conditions

Grok 4.3

Rates below are USD per 1M tokens. A threshold selects that whole input/output rate pair.

Input-token thresholdInputOutputCached input
Base, up to and including first threshold$1.25$2.5$0.2
> 200,000$2.5$5$0.4

Grok 4.6

Rates below are USD per 1M tokens. A threshold selects that whole input/output rate pair.

Input-token thresholdInputOutputCached input
Base, up to and including first threshold$2$6$0.5
> 200,000$4$12$1
Migration preflight · compare each candidate with model A

Keeping Grok 4.3 is a valid decision. No pairwise comparison certifies a drop-in replacement.

Grok 4.3 → Grok 4.6

Known loss or failed requirement

  • Context window: 1,000,000 → 500,000 tokens.
  • Output limit: 900,000 → 450,000 tokens.
  • Provider endpoint, serving precision, availability and task success need your own verification. Catalogue model capabilities do not certify an endpoint.
Check alternatives to Grok 4.3

Estimate a fixed request shape

Private to this tab unless you save locally or explicitly export private details. These are estimates, not observed invoices. Cross-model token counts and task success are assumptions.

0.1 extra attempts means 10 extra full attempts per 100 requests. Cache writes, tool charges, taxes and unlisted charges are excluded.

  • Grok 4.3Estimated total: $3.75 1,000 full attempts · base rates
  • Grok 4.6Estimated total: $7 1,000 full attempts · base rates
Prompt and output sensitivity · sampled scenarios

Every point recalculates the entire request shape. Tier-boundary samples include one token below, at and above each threshold. Ordering between samples is not guaranteed. Lowest known cost excludes unknown estimates and is not a replacement recommendation.

Vary input; hold output, requests, cache and retries fixed.
Input tokensGrok 4.3Grok 4.6Coverage
2,000$3.75Lowest known$7All computed
4,096$6.37Lowest known$11.192All computed
32,768$42.21Lowest known$68.536All computed
131,072$165.09Lowest known$265.144All computed
199,999$251.24875Lowest known$402.998All computed
200,000$251.25Lowest known$403All computed
200,001$502.5025Lowest known$806.004All computed
Vary output; hold input, requests, cache and retries fixed.
Output tokensGrok 4.3Grok 4.6
128$2.82$4.768
2,048$7.62$16.288
8,192$22.98$53.152

Bring your observed evaluation results

JSON is parsed locally. Prompts, filenames, individual responses and extra fields are not retained. Import only aggregate measurements for selected models. Task/version labels do not establish that two evaluations used the same dataset.

Expected format
[{"model":"provider/exact-model-id","task":"support-answer","version":"rubric-1","attempts":100,"successes":84,"costUsd":2.4,"latencyMs":950}]

Use the exact catalogue ID. Replace the example values with observed measurements. Omit costUsd or latencyMs when they are unknown. No public score is inferred from these observations.

Keep the decision and its evidence

A portable record retains selected facts, complete tier ladders, public requirements and the publication manifest. It also embeds the full permitted public catalogue to verify the input artifact checksum. Files are larger for this reason (up to 2 MB); local storage is limited to 12 records. Team review is local to this browser; there is no hosted collaboration.

Saving is opt-in and includes private review fields. Shared comparison links never include them. Version 1 Check references remain separate and are not overwritten or guessed into this format.

Saved local decisions · 0

No version 2 decision records are stored in this browser.

Questions this page answers

Is Grok 4.3 a better-and-cheaper replacement for Grok 4.6?

Neither model has both an equal-or-higher measured score and an equal-or-lower estimated price in this pair. Retaining your current model is a valid outcome.

Which model is higher rated: Grok 4.3 or Grok 4.6?

Grok 4.6 scores higher on LMArena general (1430.0 ±4.7) compared to Grok 4.3 (1396.8 ±3.6), a margin of 33.2 Elo points. The measured confidence intervals do not overlap, demonstrating a statistically significant preference lead on human evaluations.

Which model is more cost-effective: Grok 4.3 or Grok 4.6?

On a standard balanced workload (3 prompt tokens per 1 generated token), Grok 4.3 costs $1.56 per 1M tokens versus $3.00 for Grok 4.6. Grok 4.6 is approximately 1.9× the price of Grok 4.3. Base input rates are $1.25 for Grok 4.3 and $2.00 for Grok 4.6; base output rates are $2.50 for Grok 4.3 and $6.00 for Grok 4.6.

How do their context windows and generation limits compare?

Grok 4.3 provides a 1,000,000-token context window with a maximum output limit of 900,000 tokens. Grok 4.6 provides a 500,000-token context window with a maximum output of 450,000 tokens. Grok 4.3 offers a 2.0× larger context capacity, making it suitable for larger repositories and multi-document workflows.

What capabilities are lost when switching between Grok 4.3 and Grok 4.6?

Switching from Grok 4.3 to Grok 4.6 involves: Context window: 1,000,000 → 500,000 tokens; Output limit: 900,000 → 450,000 tokens. Run test suites on your specific prompt templates and tool declarations before cutting over production traffic.

Can Grok 4.3 or Grok 4.6 be self-hosted or deployed privately?

Both Grok 4.3 and Grok 4.6 are proprietary cloud models offered via managed API endpoints. Grok 4.3 is served by xAI and 4 catalogue offers, while Grok 4.6 is served by xAI and 4 catalogue offers.

What is included in this comparison?

Accepted model capabilities, context and output limits, complete price-tier ladders, independent LMArena measurements and missing evidence. Provider endpoint precision and application success still need verification.

Can I compare a different workload or keep a decision record?

Yes. Change the public workload, prompt length and requirements in the comparison workspace. Fixed-request scenarios, aggregate evaluations and review notes stay local unless explicitly exported.

Evidence & Ask