Claude Fable 5.1 Xhigh vs Grok 4.7 Xhigh: which scores higher?

Claude Fable 5.1 (Xhigh, Default Fallback) (Anthropic) and Grok 4.7 (Xhigh) (SpaceXAI) are not peers on the headline index, which makes the trade-offs the interesting part. The headline gap is on the intelligence index, where Claude Fable 5.1 Xhigh scores 50.0 and Grok 4.7 Xhigh 35.5. What follows is built entirely from tracked data, so where the catalogue is silent, this page says so instead of guessing.

Reasoning and overall scores

Claude Fable 5.1 Xhigh comes out ahead on the intelligence index, 50.0 to 35.5. A difference this size usually separates one tier of model from the next.

For context, that places Claude Fable 5.1 Xhigh 18th and Grok 4.7 Xhigh 38th out of 109 scored models still on offer.

Neither model has a separate coding or math index recorded yet, so those narrower comparisons will have to wait for more data.

Which is newer

The release dates are September 1, 2026 for Claude Fable 5.1 Xhigh and September 21, 2026 for Grok 4.7 Xhigh, 20 days apart. That is close enough to count them as the same generation.

Running costs

The list prices are $10.00 in and $50.00 out for Claude Fable 5.1 Xhigh, and $2.00 in and $6.00 out for Grok 4.7 Xhigh, per million tokens. Claude Fable 5.1 Xhigh costs 8.3 times as much per output token.

For a sample workload of 1 million input tokens and 200,000 output tokens, Claude Fable 5.1 Xhigh would cost $20.00 and Grok 4.7 Xhigh $3.20, a difference of $16.80. A bigger example, an agent that reads 50 million tokens and writes 5 million in a month, costs $750 on Claude Fable 5.1 Xhigh and $130 on Grok 4.7 Xhigh: a gap of $620. Blended at three input tokens per output token, the effective rates are $20.00 for Claude Fable 5.1 Xhigh and $3.00 for Grok 4.7 Xhigh.

Score per dollar

A blunt value measure, blended price over intelligence score, puts Claude Fable 5.1 Xhigh at $0.40 and Grok 4.7 Xhigh at $0.085 per point. Grok 4.7 Xhigh delivers each index point for 79% less; whether Claude Fable 5.1 Xhigh's extra 14.5 points are worth the premium depends on how often your tasks need them.

Speed and latency

Measured output speed is 61 tokens per second for Claude Fable 5.1 Xhigh and 65 tokens per second for Grok 4.7 Xhigh. Throughput is close to level. Time to first token is 98 seconds for Claude Fable 5.1 Xhigh and 43 seconds for Grok 4.7 Xhigh. Grok 4.7 Xhigh gets to its first token 2.3 times as fast. Waits that long are felt on every chat-style request, though they matter less for batch jobs.

Add the two numbers up for a 400-token reply and Grok 4.7 Xhigh finishes first: 104 seconds for Claude Fable 5.1 Xhigh, 49 seconds for Grok 4.7 Xhigh.