Claude Sonnet 5.5 vs GPT-6 Sol
Same documents, same prompt: where each model gets it right, what it really costs, and how fast it responds.
Verdict
Claude Sonnet 5.5 is more accurate; GPT-6 Sol is cheaper in 25 of 30 combinations.
More accurate
Claude Sonnet 5.5
wins 11 of 30
Cheaper
GPT-6 Sol
cheaper in 25 of 30
Faster
Claude Sonnet 5.5
2.1 sec vs 2.9 sec
Fewer fixes
Claude Sonnet 5.5
7 vs 13 documents per 1,000
Across 30 task and document-language combinations (speed and fixes: median).
At a glance
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Sonnet 5.5 | GPT-6 Sol | Difference | Claude Sonnet 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 100.00% | 99.94% | ▼ 0.06 | $9.54 | $9.65 | +1 % |
| Email triage | 98.89% | 99.04% | ▲ 0.15 | $3.21 | $2.44 | −24 % |
| Personal data detection | 100.00% | 99.71% | ▼ 0.29 | $3.22 | $2.09 | −35 % |
| Contract clauses | 99.79% | 100.00% | ▲ 0.21 | $4.23 | $3.33 | −21 % |
| Document summaries | 100.00% | 99.56% | ▼ 0.44 | $3.60 | $3.85 | +7 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $3.84 | $2.36 | −39 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Sonnet 5.5 | GPT-6 Sol | Difference | Claude Sonnet 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.95% | 99.90% | ▼ 0.05 | $10.79 | $10.12 | −6 % |
| Email triage | 98.67% | 98.96% | ▲ 0.29 | $3.92 | $3.19 | −19 % |
| Personal data detection | 99.71% | 99.71% | ±0 | $3.52 | $2.22 | −37 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $5.35 | $4.62 | −14 % |
| Document summaries | 98.67% | 100.00% | ▲ 1.33 | $5.01 | $4.72 | −6 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $5.03 | $3.95 | −21 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Sonnet 5.5 | GPT-6 Sol | Difference | Claude Sonnet 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.95% | 99.95% | ±0 | $11.21 | $10.65 | −5 % |
| Email triage | 98.00% | 98.22% | ▲ 0.22 | $4.21 | $2.73 | −35 % |
| Personal data detection | 100.00% | 99.71% | ▼ 0.29 | $3.87 | $2.13 | −45 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $6.05 | $4.36 | −28 % |
| Document summaries | 99.11% | 100.00% | ▲ 0.89 | $5.82 | $4.11 | −29 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $5.49 | $3.23 | −41 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Sonnet 5.5 | GPT-6 Sol | Difference | Claude Sonnet 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 100.00% | 99.23% | ▼ 0.77 | $10.99 | $10.79 | −2 % |
| Email triage | 97.70% | 98.44% | ▲ 0.74 | $3.76 | $2.77 | −26 % |
| Personal data detection | 100.00% | 99.81% | ▼ 0.19 | $3.78 | $2.21 | −41 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $5.70 | $4.48 | −21 % |
| Document summaries | 95.56% | 97.89% | ▲ 2.33 | $5.55 | $5.02 | −10 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $4.89 | $3.36 | −31 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Sonnet 5.5 | GPT-6 Sol | Difference | Claude Sonnet 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.95% | 99.90% | ▼ 0.05 | $10.71 | $11.00 | +3 % |
| Email triage | 97.78% | 98.59% | ▲ 0.81 | $3.85 | $2.81 | −27 % |
| Personal data detection | 100.00% | 99.71% | ▼ 0.29 | $3.81 | $2.23 | −42 % |
| Contract clauses | 99.95% | 99.59% | ▼ 0.36 | $5.71 | $4.49 | −21 % |
| Document summaries | 97.89% | 97.22% | ▼ 0.67 | $4.89 | $4.58 | −6 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $4.90 | $3.43 | −30 % |
Difference: GPT-6 Sol vs. Claude Sonnet 5.5; accuracy in percentage points, price in percent. The better value is in bold. Click a task to see the details.
Invoice extraction · English (US) (150 documents)
Claude Sonnet 5.5 is more accurate by 0.1 pp.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.94% |
| Error-free documents | 100.00% | 99.33% |
| Per 1,000 documents | $9.54 | $9.65 |
| To fix / 1,000 | 0 | 7 |
| Speed (median) | 3.5 sec | 4.1 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Invoice extraction · Czech (150 documents)
Claude Sonnet 5.5 is more accurate by 0.1 pp. GPT-6 Sol is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.95% | 99.90% |
| Error-free documents | 99.33% | 98.67% |
| Per 1,000 documents | $10.79 | $10.12 |
| To fix / 1,000 | 7 | 13 |
| Speed (median) | 4.4 sec | 3.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Invoice extraction · German (150 documents)
Both are equally accurate. GPT-6 Sol is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.95% | 99.95% |
| Error-free documents | 99.33% | 99.33% |
| Per 1,000 documents | $11.21 | $10.65 |
| To fix / 1,000 | 7 | 7 |
| Speed (median) | 6.0 sec | 4.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Invoice extraction · Spanish (150 documents)
Claude Sonnet 5.5 is more accurate by 0.8 pp.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.23% |
| Error-free documents | 100.00% | 90.00% |
| Per 1,000 documents | $10.99 | $10.79 |
| To fix / 1,000 | 0 | 100 |
| Speed (median) | 5.7 sec | 5.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Invoice extraction · French (150 documents)
Claude Sonnet 5.5 is more accurate by 0.1 pp.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.95% | 99.90% |
| Error-free documents | 99.33% | 98.67% |
| Per 1,000 documents | $10.71 | $11.00 |
| To fix / 1,000 | 7 | 13 |
| Speed (median) | 4.9 sec | 6.0 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Email triage · English (US) (150 documents)
GPT-6 Sol is more accurate by 0.1 pp. GPT-6 Sol is 1.3× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.89% | 99.04% |
| Error-free documents | 90.00% | 92.00% |
| Per 1,000 documents | $3.21 | $2.44 |
| To fix / 1,000 | 100 | 80 |
| Speed (median) | 2.0 sec | 2.8 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Email triage · Czech (150 documents)
GPT-6 Sol is more accurate by 0.3 pp. GPT-6 Sol is 1.2× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.67% | 98.96% |
| Error-free documents | 88.00% | 90.67% |
| Per 1,000 documents | $3.92 | $3.19 |
| To fix / 1,000 | 120 | 93 |
| Speed (median) | 2.0 sec | 2.8 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Email triage · German (150 documents)
GPT-6 Sol is more accurate by 0.2 pp. GPT-6 Sol is 1.5× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.00% | 98.22% |
| Error-free documents | 83.33% | 84.00% |
| Per 1,000 documents | $4.21 | $2.73 |
| To fix / 1,000 | 167 | 160 |
| Speed (median) | 1.8 sec | 2.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Email triage · Spanish (150 documents)
GPT-6 Sol is more accurate by 0.7 pp. GPT-6 Sol is 1.4× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.70% | 98.44% |
| Error-free documents | 80.67% | 86.00% |
| Per 1,000 documents | $3.76 | $2.77 |
| To fix / 1,000 | 193 | 140 |
| Speed (median) | 1.8 sec | 3.3 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Email triage · French (150 documents)
GPT-6 Sol is more accurate by 0.8 pp. GPT-6 Sol is 1.4× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.78% | 98.59% |
| Error-free documents | 82.00% | 87.33% |
| Per 1,000 documents | $3.85 | $2.81 |
| To fix / 1,000 | 180 | 127 |
| Speed (median) | 1.7 sec | 3.2 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Personal data detection · English (US) (150 documents)
Claude Sonnet 5.5 is more accurate by 0.3 pp. GPT-6 Sol is 1.5× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.71% |
| Error-free documents | 100.00% | 98.00% |
| Per 1,000 documents | $3.22 | $2.09 |
| To fix / 1,000 | 0 | 20 |
| Speed (median) | 1.9 sec | 2.6 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Personal data detection · Czech (150 documents)
Both are equally accurate. GPT-6 Sol is 1.6× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.71% | 99.71% |
| Error-free documents | 98.00% | 98.00% |
| Per 1,000 documents | $3.52 | $2.22 |
| To fix / 1,000 | 20 | 20 |
| Speed (median) | 2.2 sec | 2.5 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Personal data detection · German (150 documents)
Claude Sonnet 5.5 is more accurate by 0.3 pp. GPT-6 Sol is 1.8× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.71% |
| Error-free documents | 100.00% | 98.00% |
| Per 1,000 documents | $3.87 | $2.13 |
| To fix / 1,000 | 0 | 20 |
| Speed (median) | 1.9 sec | 2.6 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Personal data detection · Spanish (150 documents)
Claude Sonnet 5.5 is more accurate by 0.2 pp. GPT-6 Sol is 1.7× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.81% |
| Error-free documents | 100.00% | 98.67% |
| Per 1,000 documents | $3.78 | $2.21 |
| To fix / 1,000 | 0 | 13 |
| Speed (median) | 1.9 sec | 2.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Personal data detection · French (150 documents)
Claude Sonnet 5.5 is more accurate by 0.3 pp. GPT-6 Sol is 1.7× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.71% |
| Error-free documents | 100.00% | 98.00% |
| Per 1,000 documents | $3.81 | $2.23 |
| To fix / 1,000 | 0 | 20 |
| Speed (median) | 2.0 sec | 2.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Contract clauses · English (US) (150 documents)
GPT-6 Sol is more accurate by 0.2 pp. GPT-6 Sol is 1.3× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.79% | 100.00% |
| Error-free documents | 98.00% | 100.00% |
| Per 1,000 documents | $4.23 | $3.33 |
| To fix / 1,000 | 20 | 0 |
| Speed (median) | 2.1 sec | 2.5 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Contract clauses · Czech (150 documents)
Both are equally accurate. GPT-6 Sol is 1.2× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $5.35 | $4.62 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.4 sec | 2.0 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Contract clauses · German (150 documents)
Both are equally accurate. GPT-6 Sol is 1.4× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $6.05 | $4.36 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.7 sec | 2.9 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Contract clauses · Spanish (150 documents)
Both are equally accurate. GPT-6 Sol is 1.3× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $5.70 | $4.48 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.7 sec | 2.8 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Contract clauses · French (150 documents)
Claude Sonnet 5.5 is more accurate by 0.4 pp. GPT-6 Sol is 1.3× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.95% | 99.59% |
| Error-free documents | 99.33% | 94.67% |
| Per 1,000 documents | $5.71 | $4.49 |
| To fix / 1,000 | 7 | 53 |
| Speed (median) | 2.6 sec | 2.6 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Document summaries · English (US) (150 documents)
Claude Sonnet 5.5 is more accurate by 0.4 pp. Claude Sonnet 5.5 is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.56% |
| Error-free documents | 100.00% | 97.33% |
| Per 1,000 documents | $3.60 | $3.85 |
| To fix / 1,000 | 0 | 27 |
| Speed (median) | 1.8 sec | 8.1 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Document summaries · Czech (150 documents)
GPT-6 Sol is more accurate by 1.3 pp. GPT-6 Sol is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.67% | 100.00% |
| Error-free documents | 92.00% | 100.00% |
| Per 1,000 documents | $5.01 | $4.72 |
| To fix / 1,000 | 80 | 0 |
| Speed (median) | 2.7 sec | 7.6 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Document summaries · German (150 documents)
GPT-6 Sol is more accurate by 0.9 pp. GPT-6 Sol is 1.4× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.11% | 100.00% |
| Error-free documents | 94.67% | 100.00% |
| Per 1,000 documents | $5.82 | $4.11 |
| To fix / 1,000 | 53 | 0 |
| Speed (median) | 2.8 sec | 7.1 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Document summaries · Spanish (150 documents)
GPT-6 Sol is more accurate by 2.3 pp. GPT-6 Sol is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 95.56% | 97.89% |
| Error-free documents | 73.33% | 87.33% |
| Per 1,000 documents | $5.55 | $5.02 |
| To fix / 1,000 | 267 | 127 |
| Speed (median) | 2.1 sec | 7.0 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Document summaries · French (150 documents)
Claude Sonnet 5.5 is more accurate by 0.7 pp. GPT-6 Sol is 1.1× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.89% | 97.22% |
| Error-free documents | 87.33% | 83.33% |
| Per 1,000 documents | $4.89 | $4.58 |
| To fix / 1,000 | 127 | 167 |
| Speed (median) | 2.7 sec | 6.1 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Questions about a document · English (US) (150 documents)
Both are equally accurate. GPT-6 Sol is 1.6× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $3.84 | $2.36 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.5 sec | 2.3 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Questions about a document · Czech (150 documents)
Both are equally accurate. GPT-6 Sol is 1.3× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $5.03 | $3.95 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.1 sec | 2.3 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Questions about a document · German (150 documents)
Both are equally accurate. GPT-6 Sol is 1.7× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $5.49 | $3.23 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.9 sec | 2.2 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Questions about a document · Spanish (150 documents)
Both are equally accurate. GPT-6 Sol is 1.5× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $4.89 | $3.36 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.4 sec | 2.5 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.
Questions about a document · French (150 documents)
Both are equally accurate. GPT-6 Sol is 1.4× cheaper.
| Metric | Claude Sonnet 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $4.90 | $3.43 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.3 sec | 2.5 sec |
| Valid answers | 100.00% | 100.00% |
Free with login
Accuracy field by field and the documents where the models differ.
Just your email, no password.