Models/xAI/Grok 4.5

Grok 4.5

Previous Grok 4.x. 236B and #43 on /models Top Weekly. AA 56 at high. OpenRouter $2 / $6. Retired from the Continuum hosted lineup on 17 Aug 2026. Hosted xAI is grok-4.6. No official DeepSWE or vals row.

01

Identity

500,000Context in
500,000Context out
2026-07-08 (OpenRouter created 1783523154)Released
ProprietaryWeights
Canonical name
Grok 4.5
Aliases
grok-4.5, x-ai/grok-4.5
Continuum hosted id
Not listed as a Continuum hosted id
OpenRouter slug
x-ai/grok-4.5
Hugging Face
None (closed or unpublished)
Modalities
text
Closed weights Not on Continuum host list Vendor lab

Closed. OpenRouter catalog 19 Aug 2026: $2 / $6, cache read $0.30 /M, web search $0.005/request, 500k / 500k, created unix 1783523154.

HostedKeyProvider retired grok-4.5 from the hosted lineup on 2026-08-17. Do not read an older Grok 4.6 line that still said Continuum hosts 4.5.

02

Should I use this for coding agents

Use Grok 4.5 only when you specifically want this previous SKU. AA Intelligence Index 56 at high. Current flagship is Grok 4.6 (AA 58, official DeepSWE on that card, hosted).

DeepSWE

No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.

omitted, not invented as of 2026-08-13 source
Artificial Analysispublished integer
56
bench
Artificial Analysis
version
Intelligence Index
harness
published integer
metric
Intelligence Index
value
56
independent as of 2026-08-19 source
Vals Index

We did not open a vals.ai card for this identity. Chip omitted.

omitted, not invented as of 2026-08-19 source

Ranked view: Best coding models: the independent leaderboard puts this row and every other card that carries an independent coding score on one board, with the confidence intervals left visible.

Named benches, printed only

Copied from AA or vals HTML we opened. 0.0% placeholder bars are omitted. Do not average these into the hero tiles, and do not invent CI, $/task, or steps for Artificial Analysis.

BenchPrintedNoteSourceURLAs of
AA throughput 51 tok/s Printed Speed row. Artificial Analysis Grok 4.5 source 2026-08-19
AA Index eval cost $579.21 Printed Index-eval total. Artificial Analysis Grok 4.5 source 2026-08-19
03

When not to use it

Trust this before you buy

  • Need the current Grok. That is grok-4.6, hosted, with the DeepSWE row on its own card.
  • Need hosted Continuum inference. 4.5 was retired from the hosted lineup on 17 Aug 2026.
  • Need DeepSWE or vals Accuracy. Neither board printed a 4.5 row in this cluster. AA 404 is not the issue. AA 56 exists; the other two boards do not.

Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.

04

Artifact

Weights

xAI does not publish Grok 4.5 weights. No HF Files button.

  • OpenRouter catalog 19 Aug 2026: $2 / $6, cache read $0.30 /M, web search $0.005 per call. AA prints the same $2 / $6 pair, 51 tok/s, $579.21 Index eval.
  • Continuum hosted lineup: grok-4.6 only after 2026-08-17.

What you can open

Lab card and OpenRouter listing only. No Hugging Face Files button, because there is no official repo to point at.

Fetched OpenRouter catalog on 2026-08-19. hugging_face_id was empty.

05

Continuum serving

Not listed as a Continuum hosted id. Retired 17 Aug 2026. Pair a local xAI key or OpenRouter x-ai/grok-4.5.

Continuum hosted id not on the host list

We list first-party and OpenRouter, plus Continuum only when the live public allowlist named the id. This is not a 15-host routing table.

06

Economics

No DeepSWE $/task.

OpenRouter + AA, 19 Aug 2026: $2 / $6 per million. Cache read $0.30 /M on the catalog. Search $0.005/request.

Price this model Opens the pricing calculator preloaded with Grok 4.5.

Artificial Analysis Intelligence Index 56 at high, fetched 19 Aug 2026. Same page: 51 tok/s; $579.21 Index eval. The tile is the published integer 56.

07

Provenance

ClaimSourceAs of
OR catalog $2 / $6, cache $0.30, search $0.005, 500k, created 1783523154OpenRouter catalog2026-08-19
AA Index 56 (high); $2 / $6; 51 tok/s; $579.21 Index evalArtificial Analysis Grok 4.52026-08-19
Hosted grok-4.5 retired 2026-08-17Continuum HostedKeyProvider2026-08-19
236B week tokens, #43 Top WeeklyOpenRouter rankings Top Weekly2026-08-19
OpenRouter id x-ai/grok-4.5, context 500,000OpenRouter /api/v1/models2026-08-19
08

Compare, FAQ, and Get Plus

No other card in this cluster shares a published DeepSWE official row we can put next to this one. We will not compare on vendor-blog numbers.

Continue through xAI's model family with Grok Build 0.1.

FAQ

Does Continuum still host Grok 4.5?

No. The bundled catalog comment dated 2026-08-17 retires grok-4.5. The live hosted id is grok-4.6.

Why does the lab blog disagree with DeepSWE or Scale?

Lab posts pick a harness, an effort, a split, and sometimes a private eval set. DeepSWE publishes the official mini-swe-agent row with cost, tokens, and steps. Scale SWE-bench Pro only counts when the public shared-harness board has a row we can fetch. A higher lab number is usually a different test, not a better one.

OpenAI-compatible call

No Continuum hosted id is verified for this model, so this page does not invent a curl target. Use the first-party API or the OpenRouter slug x-ai/grok-4.5.

Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.