Is grok-composer-2.5-fast cheap? What 0.2x of $3/$15 actually costs

The same multiplier is not the same price

grok-composer-2.5-fast is billed at 0.2x, and so is every other Grok id in the catalog — text, image and video alike. Multipliers across the catalog run from 0.1x to 0.5x, so 0.2x is unremarkable, and because the whole family shares it, the multiplier alone makes Composer look interchangeable with its siblings.

The multiplier is a discount, and a discount is only as good as the rate it applies to. Composer's official rate is $3 per million input tokens and $15 per million output, the highest of any Grok text id — grok-4.7, grok-4.6 and grok-4.5 all list $2/$6. At 0.2x that becomes $0.60 per million input and $3.00 per million output.

Compare that with gemini-3.8-flash, which carries the same 0.2x multiplier on a $0.75/$3.75 official rate and lands at $0.15 and $0.75 per million. Same discount, and Composer costs four times as much on input and four times as much on output. Two ids with identical multipliers can sit a factor of four apart on the bill.

Against the other Grok ids, there is no break-even

Within the Grok text line there is no traffic shape where Composer comes out ahead on price. grok-4.6 and grok-4.5 both cost $0.40 per million input and $1.20 per million output. Composer is dearer on both lines: $0.60 on input, 1.5 times as much, and $3.00 on output, 2.5 times as much.

When one id costs more on input and more on output, no input-to-output ratio can rescue it, so there is no break-even point to look for. A 100,000-token prompt returning 2,000 tokens costs $0.0660 on Composer against $0.0424 on grok-4.6. A 5,000-token prompt returning 8,000 tokens costs $0.0270 against $0.0116. Input-heavy or output-heavy, only the size of the gap changes: it sits between 1.5 and 2.5 times, closer to 2.5 the more of the bill is output.

So the question is not which id suits your traffic shape — on price alone, Composer is the dearer one at every shape. The question is whether what it offers instead is worth up to 2.5 times the bill, which the next section takes up. One caveat on the arithmetic: the $0.40/$1.20 figures for grok-4.6 and grok-4.5 are the sub-200K tier, and xAI doubles both at or above 200,000 tokens in a single prompt, to $0.80/$2.40 — which is also the point where Composer's own window runs out.

curl https://token-share.app/v1/chat/completions \
  -H "Authorization: Bearer $TOKEN_SHARE_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "grok-composer-2.5-fast",
    "max_tokens": 2048,
    "messages": [
      {"role": "user", "content": "Apply this patch and report any conflicts."}
    ]
  }'

What the id gives up, and what it is for

Composer lists text and tool-calling. It does not list thinking, which grok-4.6 and grok-4.5 both do. If your prompts depend on an extended reasoning step, this is a hard difference rather than a quality judgement — the capability is simply not on the id.

The window is smaller as well: 200,000 tokens of context against 500,000 on grok-4.6 and grok-4.5, and a 32,768-token output ceiling against 65,536. A prompt that fits on 4.6 may not fit here, and a response that fits there may be cut in half.

On the numbers in the table, then, Composer costs more and offers less. What the catalog says it is for is speed: its description calls it a low-latency agentic coding tier, optimized for interactive long-running software engineering workflows. That is its whole case — a loop where someone is waiting on every turn and response time matters more than the per-token rate. The catalog publishes no latency figure behind that description, so measure it: time the same agent turn on Composer and on grok-4.6, and decide whether the difference is worth 1.5 times the input rate and 2.5 times the output rate.

Common questions