GPT-6 Sol vs Mistral Small 4
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; Mistral Small 4 is cheaper in 30 of 30 combinations.
More accurate
GPT-6 Sol
wins 23 of 30
Cheaper
Mistral Small 4
cheaper in 30 of 30
Faster
Mistral Small 4
1.0 sec vs 2.9 sec
Fewer fixes
GPT-6 Sol
13 vs 100 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 | GPT-6 Sol | Mistral Small 4 | Difference | GPT-6 Sol | Mistral Small 4 | Difference |
| Invoice extraction | 99.94% | 95.58% | ▼ 4.36 | $9.65 | $0.41 | −96 % |
| Email triage | 99.04% | 89.70% | ▼ 9.34 | $2.44 | $0.09 | −96 % |
| Personal data detection | 99.71% | 99.43% | ▼ 0.28 | $2.09 | $0.08 | −96 % |
| Contract clauses | 100.00% | 91.54% | ▼ 8.46 | $3.33 | $0.17 | −95 % |
| Document summaries | 99.56% | 99.67% | ▲ 0.11 | $3.85 | $0.14 | −96 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $2.36 | $0.15 | −94 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | GPT-6 Sol | Mistral Small 4 | Difference | GPT-6 Sol | Mistral Small 4 | Difference |
| Invoice extraction | 99.90% | 93.64% | ▼ 6.26 | $10.12 | $0.44 | −96 % |
| Email triage | 98.96% | 87.63% | ▼ 11.33 | $3.19 | $0.12 | −96 % |
| Personal data detection | 99.71% | 98.76% | ▼ 0.95 | $2.22 | $0.09 | −96 % |
| Contract clauses | 100.00% | 99.03% | ▼ 0.97 | $4.62 | $0.23 | −95 % |
| Document summaries | 100.00% | 99.00% | ▼ 1 | $4.72 | $0.22 | −95 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $3.95 | $0.21 | −95 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | GPT-6 Sol | Mistral Small 4 | Difference | GPT-6 Sol | Mistral Small 4 | Difference |
| Invoice extraction | 99.95% | 94.82% | ▼ 5.13 | $10.65 | $0.44 | −96 % |
| Email triage | 98.22% | 87.26% | ▼ 10.96 | $2.73 | $0.11 | −96 % |
| Personal data detection | 99.71% | 99.14% | ▼ 0.57 | $2.13 | $0.09 | −96 % |
| Contract clauses | 100.00% | 94.26% | ▼ 5.74 | $4.36 | $0.21 | −95 % |
| Document summaries | 100.00% | 99.89% | ▼ 0.11 | $4.11 | $0.19 | −95 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $3.23 | $0.18 | −95 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | GPT-6 Sol | Mistral Small 4 | Difference | GPT-6 Sol | Mistral Small 4 | Difference |
| Invoice extraction | 99.23% | 92.36% | ▼ 6.87 | $10.79 | $0.44 | −96 % |
| Email triage | 98.44% | 88.67% | ▼ 9.77 | $2.77 | $0.11 | −96 % |
| Personal data detection | 99.81% | 99.05% | ▼ 0.76 | $2.21 | $0.09 | −96 % |
| Contract clauses | 100.00% | 100.00% | ±0 | $4.48 | $0.23 | −95 % |
| Document summaries | 97.89% | 97.78% | ▼ 0.11 | $5.02 | $0.18 | −96 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $3.36 | $0.18 | −95 % |
| Accuracy (correct fields) | Per 1,000 documents | |||||
|---|---|---|---|---|---|---|
| Task · documents | GPT-6 Sol | Mistral Small 4 | Difference | GPT-6 Sol | Mistral Small 4 | Difference |
| Invoice extraction | 99.90% | 94.87% | ▼ 5.03 | $11.00 | $0.44 | −96 % |
| Email triage | 98.59% | 90.52% | ▼ 8.07 | $2.81 | $0.11 | −96 % |
| Personal data detection | 99.71% | 98.95% | ▼ 0.76 | $2.23 | $0.09 | −96 % |
| Contract clauses | 99.59% | 99.23% | ▼ 0.36 | $4.49 | $0.23 | −95 % |
| Document summaries | 97.22% | 93.22% | ▼ 4 | $4.58 | $0.19 | −96 % |
| Questions about a document | 100.00% | 100.00% | ±0 | $3.43 | $0.18 | −95 % |
Difference: Mistral Small 4 vs. GPT-6 Sol; 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)
GPT-6 Sol is more accurate by 4.4 pp. Mistral Small 4 is 23.7× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.94% | 95.58% |
| Error-free documents | 99.33% | 58.00% |
| Per 1,000 documents | $9.65 | $0.41 |
| To fix / 1,000 | 7 | 420 |
| Speed (median) | 4.1 sec | 1.8 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.
Invoice extraction · Czech (150 documents)
GPT-6 Sol is more accurate by 6.3 pp. Mistral Small 4 is 23.1× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.90% | 93.64% |
| Error-free documents | 98.67% | 56.00% |
| Per 1,000 documents | $10.12 | $0.44 |
| To fix / 1,000 | 13 | 440 |
| Speed (median) | 3.9 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.
Invoice extraction · German (150 documents)
GPT-6 Sol is more accurate by 5.1 pp. Mistral Small 4 is 24.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.95% | 94.82% |
| Error-free documents | 99.33% | 51.33% |
| Per 1,000 documents | $10.65 | $0.44 |
| To fix / 1,000 | 7 | 487 |
| Speed (median) | 4.9 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.
Invoice extraction · Spanish (150 documents)
GPT-6 Sol is more accurate by 6.9 pp. Mistral Small 4 is 24.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.23% | 92.36% |
| Error-free documents | 90.00% | 32.67% |
| Per 1,000 documents | $10.79 | $0.44 |
| To fix / 1,000 | 100 | 673 |
| Speed (median) | 5.9 sec | 2.1 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 · French (150 documents)
GPT-6 Sol is more accurate by 5 pp. Mistral Small 4 is 24.8× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.90% | 94.87% |
| Error-free documents | 98.67% | 52.67% |
| Per 1,000 documents | $11.00 | $0.44 |
| To fix / 1,000 | 13 | 473 |
| Speed (median) | 6.0 sec | 2.1 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.
Email triage · English (US) (150 documents)
GPT-6 Sol is more accurate by 9.3 pp. Mistral Small 4 is 25.7× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.04% | 89.70% |
| Error-free documents | 92.00% | 37.33% |
| Per 1,000 documents | $2.44 | $0.09 |
| To fix / 1,000 | 80 | 627 |
| Speed (median) | 2.8 sec | 0.8 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 11.3 pp. Mistral Small 4 is 27× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 98.96% | 87.63% |
| Error-free documents | 90.67% | 31.33% |
| Per 1,000 documents | $3.19 | $0.12 |
| To fix / 1,000 | 93 | 687 |
| Speed (median) | 2.8 sec | 0.8 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 11 pp. Mistral Small 4 is 25× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 98.22% | 87.26% |
| Error-free documents | 84.00% | 29.33% |
| Per 1,000 documents | $2.73 | $0.11 |
| To fix / 1,000 | 160 | 707 |
| Speed (median) | 2.9 sec | 0.8 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 · Spanish (150 documents)
GPT-6 Sol is more accurate by 9.8 pp. Mistral Small 4 is 25.1× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 98.44% | 88.67% |
| Error-free documents | 86.00% | 35.33% |
| Per 1,000 documents | $2.77 | $0.11 |
| To fix / 1,000 | 140 | 647 |
| Speed (median) | 3.3 sec | 0.9 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 · French (150 documents)
GPT-6 Sol is more accurate by 8.1 pp. Mistral Small 4 is 24.7× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 98.59% | 90.52% |
| Error-free documents | 87.33% | 44.67% |
| Per 1,000 documents | $2.81 | $0.11 |
| To fix / 1,000 | 127 | 553 |
| Speed (median) | 3.2 sec | 0.9 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 · English (US) (150 documents)
GPT-6 Sol is more accurate by 0.3 pp. Mistral Small 4 is 27.1× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.71% | 99.43% |
| Error-free documents | 98.00% | 96.00% |
| Per 1,000 documents | $2.09 | $0.08 |
| To fix / 1,000 | 20 | 40 |
| Speed (median) | 2.6 sec | 0.8 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.
Personal data detection · Czech (150 documents)
GPT-6 Sol is more accurate by 0.9 pp. Mistral Small 4 is 24× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.71% | 98.76% |
| Error-free documents | 98.00% | 92.00% |
| Per 1,000 documents | $2.22 | $0.09 |
| To fix / 1,000 | 20 | 80 |
| Speed (median) | 2.5 sec | 0.9 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 · German (150 documents)
GPT-6 Sol is more accurate by 0.6 pp. Mistral Small 4 is 23.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.71% | 99.14% |
| Error-free documents | 98.00% | 94.00% |
| Per 1,000 documents | $2.13 | $0.09 |
| To fix / 1,000 | 20 | 60 |
| Speed (median) | 2.6 sec | 0.9 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 · Spanish (150 documents)
GPT-6 Sol is more accurate by 0.8 pp. Mistral Small 4 is 23.6× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.81% | 99.05% |
| Error-free documents | 98.67% | 93.33% |
| Per 1,000 documents | $2.21 | $0.09 |
| To fix / 1,000 | 13 | 67 |
| Speed (median) | 2.9 sec | 0.9 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 0.8 pp. Mistral Small 4 is 23.5× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.71% | 98.95% |
| Error-free documents | 98.00% | 93.33% |
| Per 1,000 documents | $2.23 | $0.09 |
| To fix / 1,000 | 20 | 67 |
| Speed (median) | 2.9 sec | 1.0 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 · English (US) (150 documents)
GPT-6 Sol is more accurate by 8.5 pp. Mistral Small 4 is 19.6× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 91.54% |
| Error-free documents | 100.00% | 43.33% |
| Per 1,000 documents | $3.33 | $0.17 |
| To fix / 1,000 | 0 | 567 |
| Speed (median) | 2.5 sec | 1.1 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.
Contract clauses · Czech (150 documents)
GPT-6 Sol is more accurate by 1 pp. Mistral Small 4 is 20.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 99.03% |
| Error-free documents | 100.00% | 92.00% |
| Per 1,000 documents | $4.62 | $0.23 |
| To fix / 1,000 | 0 | 80 |
| Speed (median) | 2.0 sec | 1.2 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.
Contract clauses · German (150 documents)
GPT-6 Sol is more accurate by 5.7 pp. Mistral Small 4 is 21× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 94.26% |
| Error-free documents | 100.00% | 46.67% |
| Per 1,000 documents | $4.36 | $0.21 |
| To fix / 1,000 | 0 | 533 |
| Speed (median) | 2.9 sec | 1.2 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.
Contract clauses · Spanish (150 documents)
Both are equally accurate. Mistral Small 4 is 19.2× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $4.48 | $0.23 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.8 sec | 1.3 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.
Contract clauses · French (150 documents)
GPT-6 Sol is more accurate by 0.4 pp. Mistral Small 4 is 19.3× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.59% | 99.23% |
| Error-free documents | 94.67% | 90.00% |
| Per 1,000 documents | $4.49 | $0.23 |
| To fix / 1,000 | 53 | 100 |
| Speed (median) | 2.6 sec | 1.2 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.
Document summaries · English (US) (150 documents)
Mistral Small 4 is more accurate by 0.1 pp. Mistral Small 4 is 27× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 99.56% | 99.67% |
| Error-free documents | 97.33% | 98.67% |
| Per 1,000 documents | $3.85 | $0.14 |
| To fix / 1,000 | 27 | 13 |
| Speed (median) | 8.1 sec | 1.0 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 · Czech (150 documents)
GPT-6 Sol is more accurate by 1 pp. Mistral Small 4 is 21.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 99.00% |
| Error-free documents | 100.00% | 94.00% |
| Per 1,000 documents | $4.72 | $0.22 |
| To fix / 1,000 | 0 | 60 |
| Speed (median) | 7.6 sec | 1.6 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 0.1 pp. Mistral Small 4 is 21.9× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 99.89% |
| Error-free documents | 100.00% | 99.33% |
| Per 1,000 documents | $4.11 | $0.19 |
| To fix / 1,000 | 0 | 7 |
| Speed (median) | 7.1 sec | 1.3 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.
Document summaries · Spanish (150 documents)
GPT-6 Sol is more accurate by 0.1 pp. Mistral Small 4 is 27.5× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 97.89% | 97.78% |
| Error-free documents | 87.33% | 86.67% |
| Per 1,000 documents | $5.02 | $0.18 |
| To fix / 1,000 | 127 | 133 |
| Speed (median) | 7.0 sec | 1.2 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.
Document summaries · French (150 documents)
GPT-6 Sol is more accurate by 4 pp. Mistral Small 4 is 24.1× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 97.22% | 93.22% |
| Error-free documents | 83.33% | 62.00% |
| Per 1,000 documents | $4.58 | $0.19 |
| To fix / 1,000 | 167 | 380 |
| Speed (median) | 6.1 sec | 1.3 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 · English (US) (150 documents)
Both are equally accurate. Mistral Small 4 is 15.5× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $2.36 | $0.15 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.3 sec | 0.8 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.
Questions about a document · Czech (150 documents)
Both are equally accurate. Mistral Small 4 is 18.4× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $3.95 | $0.21 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.3 sec | 0.8 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.
Questions about a document · German (150 documents)
Both are equally accurate. Mistral Small 4 is 18.2× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $3.23 | $0.18 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.2 sec | 0.8 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.
Questions about a document · Spanish (150 documents)
Both are equally accurate. Mistral Small 4 is 18.3× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $3.36 | $0.18 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.5 sec | 0.8 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.
Questions about a document · French (150 documents)
Both are equally accurate. Mistral Small 4 is 18.8× cheaper.
| Metric | GPT-6 Sol | Mistral Small 4 |
|---|---|---|
| Correct fields | 100.00% | 100.00% |
| Error-free documents | 100.00% | 100.00% |
| Per 1,000 documents | $3.43 | $0.18 |
| To fix / 1,000 | 0 | 0 |
| Speed (median) | 2.5 sec | 0.8 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.