feat(pricing): add xai/grok-4.5 model pricing and metadata

[devin-ai-integration]

Relevant issues

Linear ticket

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

• I have added meaningful tests
• My PR passes all CI/CD checks (e.g., lint, format, unit tests)
• My PR's scope is as isolated as possible; it only solves 1 specific problem
• I have received a Greptile Confidence Score of at least 4/5 before requesting a maintainer review (Greptile reviews automatically once the PR is opened; only comment @greptileai to re-request a review after pushing changes)

Screenshots / Proof of Fix

xAI just shipped Grok 4.5 (grok-4.5, with grok-4.5-latest as an alias). This adds both to the cost map so cost tracking and context-window metadata resolve for the model

Pricing and metadata are taken from https://docs.x.ai/docs/models. The docs express per-token prices in internal units where the existing grok-4.3 entry (docs value 12500) maps to 1.25e-06, i.e. units x 1e-10. Applying that to grok-4.5's docs values gives input 2e-06, output 6e-06, cached input 5e-07, and the >200k long-context tier at input 4e-06, output 1.2e-05, cached 1e-06, with a 500k context window and reasoning, vision, function calling, structured outputs, prompt caching and web search all supported

Live billed call against the real xAI API, captured at commit fbad6ad8ad:

$ LITELLM_LOCAL_MODEL_COST_MAP=True XAI_API_KEY=*** python3 -c "
import litellm
from litellm import completion_cost
r=litellm.completion(model='xai/grok-4.5',messages=[{'role':'user','content':'Reply with exactly: grok-4.5 live test ok'}],max_tokens=50)
print('CONTENT:',r.choices[0].message.content)
print('MODEL:',r.model)
print('USAGE:',r.usage)
print('COST_USD:',completion_cost(completion_response=r))
"
CONTENT: grok-4.5 live test ok
MODEL: grok-4.5
USAGE: Usage(completion_tokens=59, prompt_tokens=219, total_tokens=278,
  completion_tokens_details=CompletionTokensDetailsWrapper(reasoning_tokens=50, ...),
  prompt_tokens_details=PromptTokensDetailsWrapper(cached_tokens=128, text_tokens=219, ...),
  cost_in_usd_ticks=6000000)
COST_USD: 0.0006

The $0.0006 LiteLLM computes matches the tiered pricing exactly: 91 uncached input x 2e-06 + 128 cached input x 5e-07 + 59 output x 6e-06 = 6.0e-04, which also lines up with xAI's own cost_in_usd_ticks=6000000 (ticks x 1e-10)

Metadata resolution from the local cost map:

$ LITELLM_LOCAL_MODEL_COST_MAP=True python3 -c "
import litellm
for m in ['xai/grok-4.5','xai/grok-4.5-latest']:
    i=litellm.get_model_info(m)
    print(m,'in',i['input_cost_per_token'],'out',i['output_cost_per_token'],'cache',i['cache_read_input_token_cost'],'ctx',i['max_input_tokens'],'reason',i.get('supports_reasoning'),'vision',i.get('supports_vision'),'prov',i['litellm_provider'])
"
xai/grok-4.5 in 2e-06 out 6e-06 cache 5e-07 ctx 500000 reason True vision True prov xai
xai/grok-4.5-latest in 2e-06 out 6e-06 cache 5e-07 ctx 500000 reason True vision True prov xai

Type

🆕 New Feature

Changes

Add xai/grok-4.5 and its xai/grok-4.5-latest alias to model_prices_and_context_window.json and the bundled litellm/model_prices_and_context_window_backup.json

Link to Devin session: https://app.devin.ai/sessions/e8043e650ee9475195e2caead0d4bd46
Requested by: @ishaan-berri

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[greptile-apps]

Greptile Summary

This PR adds pricing and capability metadata for xai/grok-4.5 and its xai/grok-4.5-latest alias to both the main and backup cost-map JSON files. The values have been cross-checked against the xAI public documentation and are consistent with the structure of adjacent grok-4.3 entries.

• Pricing ($2/M input, $6/M output, $0.50/M cached) and the 500k context window match the xAI docs; the above-200k tiered rates follow the same 2× multiplier pattern used by grok-4.3.
• Both model_prices_and_context_window.json and the bundled backup are updated identically, and all capability flags (supports_reasoning, supports_vision, supports_prompt_caching, etc.) align with the model's published feature set.

Confidence Score: 5/5

Safe to merge — two JSON pricing files updated with well-sourced, structurally consistent metadata.

The change is additive-only (two new JSON objects), the per-token prices match the xAI public documentation, the above-200k tier follows the same doubling pattern as grok-4.3, and both files are kept in sync. No logic, tests, or existing entries are touched.

No files require special attention.

Important Files Changed

Filename	Overview
model_prices_and_context_window.json	Adds xai/grok-4.5 and xai/grok-4.5-latest with pricing and capability metadata; values match xAI docs and are structurally consistent with existing xAI model entries.
litellm/model_prices_and_context_window_backup.json	Backup JSON updated in sync with the main pricing file; identical entries for xai/grok-4.5 and xai/grok-4.5-latest.

_{Reviews (1): Last reviewed commit: "feat(pricing): add xai/grok-4.5 model pr..." | Re-trigger Greptile}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[codspeed-hq]

Merging this PR will not alter performance

✅ 30 untouched benchmarks

---

_{Comparing litellm_add_grok_4_5 (fbad6ad) with litellm_internal_staging (bd23c44)}

[mateo-berri]

LGTM; thanks!