As of August 20, 2026, Llama 3.2 90B Instruct is 91% cheaper per million tokens than Grok 4.6. Grok 4.6 is $2.00 / $6.00 per million input/output tokens versus Llama 3.2 90B Instruct at $0.35 / $0.40. Sources: https://docs.x.ai/docs/models and https://llama.meta.com/
Sources: xAI API pricing (https://docs.x.ai/docs/models); Meta pricing (https://llama.meta.com/)
Grok 4.6
xAI
Input
$2.00
Output
$6.00
Llama 3.2 90B Instruct
Meta
Input
$0.35
Output
$0.40
Token pricing, specs, and benchmarks side by side
Grok 4.6Llama 3.2 90B InstructVerdict
Both models are closely matched on benchmarks. Llama 3.2 90B Instruct has the pricing advantage, making it the better value unless Grok 4.6's specific capabilities are critical to your workflow.
As of August 20, 2026, Llama 3.2 90B Instruct is 91% cheaper on blended LLM API pricing. Llama 3.2 90B Instruct costs $0.35 per million input tokens and $0.40 per million output tokens ($0.75 blended 1M-in + 1M-out). Grok 4.6 costs $2.00 / $6.00 per million tokens ($8.00 blended). Sources: https://docs.x.ai/docs/models and https://llama.meta.com/. The cheapest LLM for your app still depends on how many output tokens you generate.
Grok 4.6 xAI API pricing is $2.00 input and $6.00 output per million tokens via xAI. Llama 3.2 90B Instruct Meta API pricing is $0.35 input and $0.40 output per million tokens via Meta. Use those four numbers, not ChatGPT Plus or Claude Pro subscription prices, when you are comparing APIs.
Grok 4.6 has a published SWE-bench Verified score of 95.6%. Llama 3.2 90B Instruct does not have that eval on this page yet.
Grok 4.6 supports a 500K token context window, while Llama 3.2 90B Instruct supports 128K tokens. Grok 4.6 offers a larger context window. Filling a larger window bills more input tokens, so the cheaper-per-million model can still cost more on long documents.
Shared benchmark scores for Grok 4.6 and Llama 3.2 90B Instruct are still limited. Check the comparison table for the evals each model has published. Check the comparison table for GPQA Diamond, SWE-bench, MMLU, HLE, and the other evals we track. Token price and benchmark score together are the usual LLM comparison, not either number alone.
Both Grok 4.6 and Llama 3.2 90B Instruct are production API models. For cost-sensitive production traffic, Llama 3.2 90B Instruct has better economics on the blended 1M-in + 1M-out scale. For maximum capability, weight the benchmark table for your domain. Many production stacks route cheap models for drafts and a frontier model for the hard turn, which is usually cheaper than sending everything to the expensive API.
Yes. Call Grok 4.6 through xAI and Llama 3.2 90B Instruct through Meta with separate API keys, or through a gateway that already wraps both. Keep prompts in tokens, not characters, when you estimate the invoice. AnotherWrapper templates can swap providers without rewriting auth and billing.
List rates are the latest published xAI API pricing and Meta API pricing, quoted per million tokens. Official rate sheets: https://docs.x.ai/docs/models and https://llama.meta.com/. Prompt caching, batch APIs, and committed-use discounts can change the invoice. Benchmark scores are from official publications and independent evals, updated as new numbers land.
Grok 4.6 supports up to an unspecified number of output tokens per request, while Llama 3.2 90B Instruct supports up to 128K output tokens. Output tokens are usually the expensive half of API pricing, so a higher max output cap is a capability, not a discount.
Grok 4.6 runs at approximately 4.687848222704461 tokens/second while Llama 3.2 90B Instruct runs at 24 tokens/second. Llama 3.2 90B Instruct is faster in raw throughput. Throughput (tokens per second) and time-to-first-token are why two models with the same per-million-token price can still feel different in chat and agents.
Use the model pickers at the top of this LLM comparison. Search any model in the index and the URL updates. Open Grok 4.6 or Llama 3.2 90B Instruct from the sidebar for the single-model API pricing page, or go back to the LLM pricing table to sort by cheapest blended cost.
This comparison covers API pricing, context window, throughput, and benchmark scores. Grok 4.6 costs $2.00/M input and $6.00/M output. Llama 3.2 90B Instruct costs $0.35/M input and $0.40/M output.
Rank both on the LLM leaderboard, SWE-bench, and Humanity's Last Exam.
The index
280 models across 33 providers. Search or jump to a lab — every model page stays linked here.
Next step
Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.
