Claude Haiku 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
GPT-6 Sol is more accurate; Claude Haiku 5.5 is cheaper in 30 of 30 combinations.
More accurate
GPT-6 Sol
wins 18 of 30
Cheaper
Claude Haiku 5.5
cheaper in 30 of 30
Faster
Claude Haiku 5.5
2.1 sec vs 2.9 sec
Fewer fixes
GPT-6 Sol
13 vs 33 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 Haiku 5.5 | GPT-6 Sol | Difference | Claude Haiku 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 100.00% | 99.94% | ▼ 0.06 | $0.61 | $9.65 | +1492 % |
| Email triage | 98.15% | 99.04% | ▲ 0.89 | $0.23 | $2.44 | +940 % |
| Personal data detection | 99.81% | 99.71% | ▼ 0.1 | $0.19 | $2.09 | +1014 % |
| Contract clauses | 99.69% | 100.00% | ▲ 0.31 | $0.35 | $3.33 | +838 % |
| Document summaries | 98.33% | 99.56% | ▲ 1.23 | $0.34 | $3.85 | +1033 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $0.23 | $2.36 | +924 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Haiku 5.5 | GPT-6 Sol | Difference | Claude Haiku 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.90% | 99.90% | ±0 | $0.62 | $10.12 | +1532 % |
| Email triage | 97.63% | 98.96% | ▲ 1.33 | $0.29 | $3.19 | +1008 % |
| Personal data detection | 99.43% | 99.71% | ▲ 0.28 | $0.21 | $2.22 | +957 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $0.40 | $4.62 | +1058 % |
| Document summaries | 95.89% | 100.00% | ▲ 4.11 | $0.32 | $4.72 | +1356 % |
| Questions about a document | 99.78% | 100.00% | ▲ 0.22 | $0.30 | $3.95 | +1207 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Haiku 5.5 | GPT-6 Sol | Difference | Claude Haiku 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.85% | 99.95% | ▲ 0.1 | $0.71 | $10.65 | +1404 % |
| Email triage | 97.11% | 98.22% | ▲ 1.11 | $0.32 | $2.73 | +752 % |
| Personal data detection | 99.62% | 99.71% | ▲ 0.09 | $0.23 | $2.13 | +821 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $0.48 | $4.36 | +798 % |
| Document summaries | 99.44% | 100.00% | ▲ 0.56 | $0.46 | $4.11 | +788 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $0.32 | $3.23 | +902 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Haiku 5.5 | GPT-6 Sol | Difference | Claude Haiku 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.44% | 99.23% | ▼ 0.21 | $0.73 | $10.79 | +1376 % |
| Email triage | 97.70% | 98.44% | ▲ 0.74 | $0.30 | $2.77 | +823 % |
| Personal data detection | 100.00% | 99.81% | ▼ 0.19 | $0.23 | $2.21 | +865 % |
| Contract clauses | 99.95% | 100.00% | ▲ 0.05 | $0.42 | $4.48 | +965 % |
| Document summaries | 94.11% | 97.89% | ▲ 3.78 | $0.61 | $5.02 | +719 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $0.29 | $3.36 | +1067 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | Claude Haiku 5.5 | GPT-6 Sol | Difference | Claude Haiku 5.5 | GPT-6 Sol | Difference |
| Invoice extraction | 99.85% | 99.90% | ▲ 0.05 | $0.77 | $11.00 | +1325 % |
| Email triage | 97.41% | 98.59% | ▲ 1.18 | $0.31 | $2.81 | +799 % |
| Personal data detection | 99.33% | 99.71% | ▲ 0.38 | $0.23 | $2.23 | +865 % |
| Contract clauses | 99.38% | 99.59% | ▲ 0.21 | $0.45 | $4.49 | +891 % |
| Document summaries | 92.78% | 97.22% | ▲ 4.44 | $0.66 | $4.58 | +590 % |
| Questions about a document | 99.22% | 100.00% | ▲ 0.78 | $0.28 | $3.43 | +1107 % |
Difference: GPT-6 Sol vs. Claude Haiku 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 Haiku 5.5 is more accurate by 0.1 pp. Claude Haiku 5.5 is 15.9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.94% |
| Error-free documents | 100.00% | 99.33% |
| Per 1,000 documents | $0.61 | $9.65 |
| To fix / 1,000 | 0 | 7 |
| Speed (median) | 3.4 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)
Both are equally accurate. Claude Haiku 5.5 is 16.3× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.90% | 99.90% |
| Error-free documents | 98.67% | 98.67% |
| Per 1,000 documents | $0.62 | $10.12 |
| To fix / 1,000 | 13 | 13 |
| Speed (median) | 3.4 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)
GPT-6 Sol is more accurate by 0.1 pp. Claude Haiku 5.5 is 15× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.85% | 99.95% |
| Error-free documents | 98.00% | 99.33% |
| Per 1,000 documents | $0.71 | $10.65 |
| To fix / 1,000 | 20 | 7 |
| Speed (median) | 3.9 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 Haiku 5.5 is more accurate by 0.2 pp. Claude Haiku 5.5 is 14.8× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.44% | 99.23% |
| Error-free documents | 92.67% | 90.00% |
| Per 1,000 documents | $0.73 | $10.79 |
| To fix / 1,000 | 73 | 100 |
| Speed (median) | 4.4 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)
Both are equally accurate. Claude Haiku 5.5 is 14.2× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.85% | 99.90% |
| Error-free documents | 98.00% | 98.67% |
| Per 1,000 documents | $0.77 | $11.00 |
| To fix / 1,000 | 20 | 13 |
| Speed (median) | 4.6 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 0.9 pp. Claude Haiku 5.5 is 10.4× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.15% | 99.04% |
| Error-free documents | 84.00% | 92.00% |
| Per 1,000 documents | $0.23 | $2.44 |
| To fix / 1,000 | 160 | 80 |
| Speed (median) | 1.7 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.
Email triage · Czech (150 documents)
GPT-6 Sol is more accurate by 1.3 pp. Claude Haiku 5.5 is 11.1× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.63% | 98.96% |
| Error-free documents | 80.00% | 90.67% |
| Per 1,000 documents | $0.29 | $3.19 |
| To fix / 1,000 | 200 | 93 |
| Speed (median) | 1.9 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 1.1 pp. Claude Haiku 5.5 is 8.5× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.11% | 98.22% |
| Error-free documents | 74.67% | 84.00% |
| Per 1,000 documents | $0.32 | $2.73 |
| To fix / 1,000 | 253 | 160 |
| Speed (median) | 2.1 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. Claude Haiku 5.5 is 9.2× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.70% | 98.44% |
| Error-free documents | 80.00% | 86.00% |
| Per 1,000 documents | $0.30 | $2.77 |
| To fix / 1,000 | 200 | 140 |
| Speed (median) | 2.1 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 1.2 pp. Claude Haiku 5.5 is 9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 97.41% | 98.59% |
| Error-free documents | 76.67% | 87.33% |
| Per 1,000 documents | $0.31 | $2.81 |
| To fix / 1,000 | 233 | 127 |
| Speed (median) | 2.2 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 Haiku 5.5 is more accurate by 0.1 pp. Claude Haiku 5.5 is 11.1× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.81% | 99.71% |
| Error-free documents | 98.67% | 98.00% |
| Per 1,000 documents | $0.19 | $2.09 |
| To fix / 1,000 | 13 | 20 |
| Speed (median) | 1.7 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)
GPT-6 Sol is more accurate by 0.3 pp. Claude Haiku 5.5 is 10.6× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.43% | 99.71% |
| Error-free documents | 96.00% | 98.00% |
| Per 1,000 documents | $0.21 | $2.22 |
| To fix / 1,000 | 40 | 20 |
| Speed (median) | 1.8 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)
GPT-6 Sol is more accurate by 0.1 pp. Claude Haiku 5.5 is 9.2× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.62% | 99.71% |
| Error-free documents | 97.33% | 98.00% |
| Per 1,000 documents | $0.23 | $2.13 |
| To fix / 1,000 | 27 | 20 |
| Speed (median) | 1.8 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 Haiku 5.5 is more accurate by 0.2 pp. Claude Haiku 5.5 is 9.7× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 99.81% |
| Error-free documents | 100.00% | 98.67% |
| Per 1,000 documents | $0.23 | $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)
GPT-6 Sol is more accurate by 0.4 pp. Claude Haiku 5.5 is 9.6× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.33% | 99.71% |
| Error-free documents | 95.33% | 98.00% |
| Per 1,000 documents | $0.23 | $2.23 |
| To fix / 1,000 | 47 | 20 |
| 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.
Contract clauses · English (US) (150 documents)
GPT-6 Sol is more accurate by 0.3 pp. Claude Haiku 5.5 is 9.4× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.69% | 100.00% |
| Error-free documents | 97.33% | 100.00% |
| Per 1,000 documents | $0.35 | $3.33 |
| To fix / 1,000 | 27 | 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. Claude Haiku 5.5 is 11.6× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $0.40 | $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. Claude Haiku 5.5 is 9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $0.48 | $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. Claude Haiku 5.5 is 10.6× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.95% | 100.00% |
| Error-free documents | 99.33% | 100.00% |
| Per 1,000 documents | $0.42 | $4.48 |
| To fix / 1,000 | 7 | 0 |
| Speed (median) | 2.3 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)
GPT-6 Sol is more accurate by 0.2 pp. Claude Haiku 5.5 is 9.9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.38% | 99.59% |
| Error-free documents | 92.00% | 94.67% |
| Per 1,000 documents | $0.45 | $4.49 |
| To fix / 1,000 | 80 | 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)
GPT-6 Sol is more accurate by 1.2 pp. Claude Haiku 5.5 is 11.3× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 98.33% | 99.56% |
| Error-free documents | 90.00% | 97.33% |
| Per 1,000 documents | $0.34 | $3.85 |
| To fix / 1,000 | 100 | 27 |
| Speed (median) | 2.7 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 4.1 pp. Claude Haiku 5.5 is 14.6× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 95.89% | 100.00% |
| Error-free documents | 77.33% | 100.00% |
| Per 1,000 documents | $0.32 | $4.72 |
| To fix / 1,000 | 227 | 0 |
| Speed (median) | 2.1 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 0.6 pp. Claude Haiku 5.5 is 8.9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.44% | 100.00% |
| Error-free documents | 96.67% | 100.00% |
| Per 1,000 documents | $0.46 | $4.11 |
| To fix / 1,000 | 33 | 0 |
| Speed (median) | 3.5 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 3.8 pp. Claude Haiku 5.5 is 8.2× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 94.11% | 97.89% |
| Error-free documents | 64.67% | 87.33% |
| Per 1,000 documents | $0.61 | $5.02 |
| To fix / 1,000 | 353 | 127 |
| Speed (median) | 3.8 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)
GPT-6 Sol is more accurate by 4.4 pp. Claude Haiku 5.5 is 6.9× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 92.78% | 97.22% |
| Error-free documents | 56.67% | 83.33% |
| Per 1,000 documents | $0.66 | $4.58 |
| To fix / 1,000 | 433 | 167 |
| Speed (median) | 4.0 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. Claude Haiku 5.5 is 10.2× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $0.23 | $2.36 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.8 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)
GPT-6 Sol is more accurate by 0.2 pp. Claude Haiku 5.5 is 13.1× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.78% | 100.00% |
| Error-free documents | 98.67% | 100.00% |
| Per 1,000 documents | $0.30 | $3.95 |
| To fix / 1,000 | 13 | 0 |
| Speed (median) | 1.9 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. Claude Haiku 5.5 is 10× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $0.32 | $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. Claude Haiku 5.5 is 11.7× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $0.29 | $3.36 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 1.8 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)
GPT-6 Sol is more accurate by 0.8 pp. Claude Haiku 5.5 is 12.1× cheaper.
| Metric | Claude Haiku 5.5 | GPT-6 Sol |
|---|---|---|
| Correct fields | 99.22% | 100.00% |
| Error-free documents | 96.67% | 100.00% |
| Per 1,000 documents | $0.28 | $3.43 |
| To fix / 1,000 | 33 | 0 |
| Speed (median) | 1.8 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.