Is Claude Sonnet 4 a good deal?

Whether anything in this catalogue beats Claude Sonnet 4 on both quality and price, and what you give up if it does. A computation on the current catalogue, not an opinion.

13 undominated of 136 · Sep 23, 2026

As of Sep 23, 2026, Claude Sonnet 4 is dominated for Balanced on LMArena. Gemini 3.8 Flash scores 155.4 higher and costs 75% less, with a covering envelope. 13 of 136 rated, priced standard models are undominated.

Inspect model evidence Compare differences & requirements

Current model

Claude Sonnet 4

Balanced

3 tokens in per 1 out

Gemini 3.8 Flash is both better and cheaper than Claude Sonnet 4: 155.4 points higher on LMArena and 75% less per million tokens, $4.50 cheaper at this mix.

Claude Sonnet 4

LMArena Elo · higher is better

Scale starts at 1040 Elo

1339.3

Effective $/M · Balanced · lower is better

$6.00/M

Gemini 3.8 Flash

LMArena Elo · higher is better

Scale starts at 1040 Elo

1494.7

Effective $/M · Balanced · lower is better

$1.50/M

Claude Sonnet 4 takes image, text, file, returns up to 64,000 tokens from a 200,000-token context, and is listed by 2 sellers.

Compared against 136 rated, priced models on this lens: 32 models dominate it and give up nothing, 5 more dominate it but give something up. Claude Sonnet 4 scores 1339.3 at $6.00 per million tokens for this mix.

Envelope-safe replacements

Each row scores at least as high, costs no more, and covers this model’s context, output, modalities, tools, and reasoning. A cheaper narrower model is not listed here.

ModelLMArenaEffective $/MYou save
Gemini 3.8 Flash1494.7 +155.4$1.50/M75%
Gemini 3.7 Flash1490.5 +151.2$1.50/M75%
Muse Spark 1.31489.7 +150.4$2.00/M67%
Muse Spark 1.11480.2 +140.9$2.00/M67%
Gemini 3.1 Pro Preview1480.1 +140.8$4.50/M25%
Gemini 3.6 Flash1476.1 +136.8$1.50/M75%
Gemini 3.5 Flash1475.7 +136.4$3.38/M44%
Claude Sonnet 4.61458.3 +119.0$6.00/M0%
Gemini 2.5 Pro1457.8 +118.5$3.44/M43%
GPT-5.6 Sol1455.1 +115.8$4.00/M33%
GPT-5.41452.6 +113.3$5.63/M6%
Grok 4.51450.1 +110.8$3.00/M50%
GPT-5.6 Terra1446.2 +106.9$4.50/M25%
Claude Sonnet 51442.2 +102.9$4.00/M33%
Claude Sonnet 4.51438.3 +99.0$6.00/M0%
Gemini 3.5 Flash Lite1435.5 +96.2$0.850/M86%
GPT-5.6 Luna1429.9 +90.6$0.450/M93%
Grok 4.61429.9 +90.6$3.00/M50%
GPT-5.11422.6 +83.3$3.44/M43%
Mistral Medium 3.51420.6 +81.3$3.00/M50%
Gemini 2.5 Flash1417.3 +78.0$0.850/M86%
Gemini 3.1 Flash Lite Preview1415.3 +76.0$0.563/M91%
GPT-5.21412.4 +73.1$4.81/M20%
GPT-5.4 Mini1412.1 +72.8$1.69/M72%
o31409.9 +70.6$3.50/M42%
GPT-51406.1 +66.8$3.44/M43%
Grok 4.31397.7 +58.4$1.56/M74%
Claude Haiku 4.51396.6 +57.3$2.00/M67%
GPT-5 Mini1373.0 +33.7$0.688/M89%
GPT-5.4 Nano1372.9 +33.6$0.463/M92%
Nova 2 Lite1362.6 +23.3$0.850/M86%
o4 Mini1353.2 +13.9$1.93/M68%

Higher score, lower price, named losses

Not a drop-in. The loss is why these are not a recommendation.

ModelLMArenaEffective $/MYou give up
GLM 5.31475.1$1.29/Mimage, file
Kimi K31472.3$6.00/Mfile
GLM 5.3 Flash1471.9$0.237/Mfile
GLM 5.21466.9$0.998/Mimage, file
MiMo-V2.5-Pro1464.8$0.544/Mimage, file
Gemini 3.8 Flash1494.7 · $1.50/MClaude Sonnet 41339.3 · $6.00/MLMArena EloEffective $/M · Balanced
136 rated, priced standard models. Chartreuse is the frontier. Cobalt is the model you named, when it is not on the staircase.

Open the frontier

The link keeps the model and mix, using the current catalogue. Monthly spend and switching cost stay in this browser tab and are left out of the link.

Saved decision references

Save the model, workload, catalogue date and capability-preservation rule in this browser. Spend, switching cost and bill contents are not saved. No account or notifications.

Constraint: replacements must preserve the model’s capabilities; any losses remain named trade-offs.

Historical decisions cannot be fully replayed from saved references: past prices, scores and capability evidence are not stored.

Evidence & Ask