Get started

Claude Haiku 5.5 against Grok 4.6.

Two listings from the same committed catalog, side by side: the fields OpenRouter publishes, the boards two third parties scored, and a cost sheet that is arithmetic over the listed rates.

retrieved sha256 · 0bb53a77713f

Compare/Claude Haiku 5.5vsGrok 4.6USD per 1M tokens · one scale
Listed
1,000,000 tokens
Context
500,000 tokens
128,000 tokens
Max output
450,000 tokens
text, image, file → text
Modalities
text, image, file → text
not stated
Knowledge cutoff
not stated
145 tok/s
Throughput
62.6 tok/s
1.57s
First token
2.52s
1,286,911 requests · 10 endpoints
Timed over
2,722 requests · 4 endpoints

Listed price

USD per million tokens · one linear scale
Claude Haiku 5.5Anthropic · listed 7 Oct 2026
Input per M$0.10
Cached input per M$0.01
Output per M$0.50
Blended, 3 in to 1 out$0.20
Grok 4.6SpaceXAI · listed 12 Aug 2026
Input per M$2.00
Cached input per M$0.50
Output per M$6.00
Blended, 3 in to 1 out$3.00

Per million tokens, as OpenRouter listed them when the catalog was read: its routed rates, not a provider's own page. One linear scale across every row, so a bar twice as long is twice the price. The blended row is Vitna's arithmetic, three input tokens to one output.

Every board either stands on

23 boards
36.5 · 16 of 58
Intelligence Index
31.0 · 22 of 58
not scored
Coding Index
75.9 · 10 of 118
not scored
Agentic Index
52.5 · 06 of 94
1252 · 41 of 126
Websites · Models
1276 · 27 of 126
1276 · 25 of 120
Code · Models
1278 · 23 of 120
1313 · 15 of 120
Game development · Models
1280 · 23 of 120
1267 · 32 of 116
UI components · Models
1283 · 25 of 116
1319 · 17 of 113
3D scenes · Models
1254 · 26 of 113
1277 · 19 of 110
Data visualisation · Models
1281 · 18 of 110
1196 · 31 of 83
SVG · Models
1221 · 25 of 83
1204 · 24 of 67
ASCII art · Models
1256 · 15 of 67
not ranked
Mobile apps · Agents
1224 · 15 of 49
not ranked
Web apps · Agents
1230 · 14 of 47
not ranked
Full stack · Agents
1237 · 09 of 44
not ranked
Android native · Agents
1279 · 03 of 41
1260 · 07 of 36
pdf · Agents
1231 · 12 of 36
not ranked
Godot game development · Agents
1247 · 04 of 34
not ranked
spreadsheet · Agents
1277 · 07 of 32
not ranked
python-pptx slides · Agents
1214 · 10 of 29
not ranked
codetovideo · Agents
1166 · 17 of 26
not ranked
Agentic game development · Agents
1155 · 10 of 25
not ranked
blender3d · Agents
1121 · 11 of 25
not ranked
HTML slides · Agents
1146 · 06 of 25

Score and position on each board, from the board's own source as OpenRouter carries it. The heavier figure leads that board; a missing score is a fact about coverage, not a zero.

Vitna reads

from provider pages

Where the registry tracks a model, Vitna reads the rate from the provider page itself and keeps it as a dated record; the OpenRouter figures above are its routed rates.

Anthropic · Claude Haiku 5.5$0.10 / $0.50SpaceXAI · Grok 4.6$2.00 / $6.00Google · Grok 4.6$2.00 / $6.00AWS · Grok 4.6$2.20 / $6.60

Cost sheet

800 in · 200 out per call

Arithmetic over the listed rates and nothing else: tokens in over one million times the input rate, plus tokens out over one million times the output rate. No cache, no retries and no failure rate assumed; the Decision Lab is where cost per successful unit lives.

ModelPer callPer 1,000 callsPer 100,000 calls
Claude Haiku 5.5$0.000180$0.1800$18.00
Grok 4.6$0.002800$2.80$280

Claude Haiku 5.5 costs 94% less per call at these token counts.

Other pairs with a page
For one of your workloads

The market on these two is here. The right one for your work is not.

What this page cannot say is which is right for a particular piece of your production work, because that depends on what the workload requires and what your own evaluations measured, neither of which is public. The register weighs a candidate against a workload's own thresholds and shows what the engine would decide before anything is filed.

Sources

Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Vitna ran none of these evaluations and adjusted none of these figures.