# Vals Index / terminal bench 4 > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/vals-ai%3Avals_index%3Aterminal_bench_4. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/vals-ai%3Avals_index%3Aterminal_bench_4. A single measure of AI's potential economic impact — agentic model performance across finance, coding, legal, and tax tasks, weighted by each sector's share of U.S. GDP. Canonical URL: https://benchize.com/benchmarks/vals-ai%3Avals_index%3Aterminal_bench_4 Last updated: Wed, 21 Oct 2026 18:52:20 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. A single measure of AI's potential economic impact — agentic model performance across finance, coding, legal, and tax tasks, weighted by each sector's share of U.S. GDP. Composite indexes · Models tested: 14 · Score publishers: Vals AI ## Model results | # | Model | Score | ± | Cost | Latency | | | --- | --- | --- | --- | --- | --- | --- | | 1 | [Claude Opus 5.5](https://benchize.com/models/claude-opus-5-5-max.md)Max | 65.15% | SE ±0 pp | $13.2 | 3,810.85source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 65.152 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 13.198632 \| USD \| per_test \| Lower is better \| \| `latency` \| 3810.852 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±0 pp sourcePopulation 23 stderr 0 tokenTotals {"cache_read_tokens":1824613192,"cache_write_tokens":26310456,"input_tokens":1850938631,"output_tokens":14913437,"reasoning_tokens":10690470} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label anthropic/claude-opus-5-5 Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/anthropic~1claude-opus-5-5` | | | | | | | | 2 | [Claude Sonnet 5.5](https://benchize.com/models/claude-sonnet-5-5-max.md)Max | 64.14% | SE ±1.01 pp | $16.51 | 4,928.93source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 64.141 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 16.513006 \| USD \| per_test \| Lower is better \| \| `latency` \| 4928.927 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.01 pp sourcePopulation 23 stderr 1.01 tokenTotals {"cache_read_tokens":3801766853,"cache_write_tokens":37747686,"input_tokens":3839535071,"output_tokens":23509474,"reasoning_tokens":17732859} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label anthropic/claude-sonnet-5-5 Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/anthropic~1claude-sonnet-5-5` | | | | | | | | 3 | [GPT-6 Astra](https://benchize.com/models/gpt-6-astra-max.md)Max | 59.6% | SE ±4.4 pp | $9.58 | 2,142.08source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 59.596 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 9.584259 \| USD \| per_test \| Lower is better \| \| `latency` \| 2142.082 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±4.4 pp sourcePopulation 23 stderr 4.403 tokenTotals {"cache_read_tokens":184898940,"cache_write_tokens":8367663,"input_tokens":193308076,"output_tokens":5366501,"reasoning_tokens":3845988} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label openai/gpt-6-astra Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/openai~1gpt-6-astra` | | | | | | | | 4 | [Gemini 4 Argon](https://benchize.com/models/gemini-4-argon-high.md)High | 57.58% | SE ±2.31 pp | $17.64 | 3,214.79source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 57.576 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 17.63994 \| USD \| per_test \| Lower is better \| \| `latency` \| 3214.793 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 262144 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.31 pp sourcePopulation 23 stderr 2.314 tokenTotals {"cache_read_tokens":895783049,"cache_write_tokens":0,"input_tokens":1088824555,"output_tokens":10645668,"reasoning_tokens":6605548} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label google/gemini-4-argon Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/google~1gemini-4-argon` | | | | | | | | 5 | [GPT-6.1 Sol](https://benchize.com/models/gpt-6-1-sol-max.md)Max | 55.05% | SE ±1.82 pp | $1.72 | 2,711.67source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 55.051 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 1.72496 \| USD \| per_test \| Lower is better \| \| `latency` \| 2711.667 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.82 pp sourcePopulation 23 stderr 1.821 tokenTotals {"cache_read_tokens":172615332,"cache_write_tokens":9543063,"input_tokens":182186678,"output_tokens":6309120,"reasoning_tokens":4640611} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label openai/gpt-6.1-sol Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/openai~1gpt-6.1-sol` | | | | | | | | 6 | [Muse Wonder](https://benchize.com/models/muse-wonder-max.md)Max | 49.5% | SE ±2.2 pp | $14.66 | 3,280.69source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 49.495 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 14.657694 \| USD \| per_test \| Lower is better \| \| `latency` \| 3280.687 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 131072 provider Meta reasoning Not reported Reasoning effort max Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.2 pp sourcePopulation 23 stderr 2.201 tokenTotals {"cache_read_tokens":2198750960,"cache_write_tokens":0,"input_tokens":2276758314,"output_tokens":10781411,"reasoning_tokens":7812084} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label meta/muse-wonder Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/meta~1muse-wonder` | | | | | | | | 7 | [Claude Haiku 5.5](https://benchize.com/models/claude-haiku-5-5-max.md)Max | 36.36% | SE ±2.31 pp | $3.12 | 3,116.39source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 36.364 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 3.117464 \| USD \| per_test \| Lower is better \| \| `latency` \| 3116.385 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 128000 provider Anthropic reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.31 pp sourcePopulation 23 stderr 2.314 tokenTotals {"cache_read_tokens":2952152829,"cache_write_tokens":29466242,"input_tokens":2981637313,"output_tokens":21606148,"reasoning_tokens":16498863} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label anthropic/claude-haiku-5-5 Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/anthropic~1claude-haiku-5-5` | | | | | | | | 8 | [DeepSeek V4.1 Flash 1017](https://benchize.com/models/deepseek-v4-1-flash-1017-max.md)Max | 31.82% | SE ±2.62 pp | $0.54 | 2,470.45source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 31.818 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.539359 \| USD \| per_test \| Lower is better \| \| `latency` \| 2470.449 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 384000 provider DeepSeek reasoning Not reported Reasoning effort max Temperature 1 topP 1 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.62 pp sourcePopulation 23 stderr 2.624 tokenTotals {"cache_read_tokens":2592293631,"cache_write_tokens":0,"input_tokens":2602850452,"output_tokens":14064073,"reasoning_tokens":10223328} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label deepseek/deepseek-v4.1-flash-1017 Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/deepseek~1deepseek-v4.1-flash-1017` | | | | | | | | 9 | [MiMo-V2.6-Pro](https://benchize.com/models/mimo-v2-6-pro.md) | 31.31% | SE ±3.07 pp | $0.5 | 9,720.05source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 31.313 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.497505 \| USD \| per_test \| Lower is better \| \| `latency` \| 9720.048 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Xiaomi reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±3.07 pp sourcePopulation 23 stderr 3.072 tokenTotals {"cache_read_tokens":1715650773,"cache_write_tokens":0,"input_tokens":1740494990,"output_tokens":15117928,"reasoning_tokens":12015515} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label xiaomi/mimo-v2.6-pro Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/xiaomi~1mimo-v2.6-pro` | | | | | | | | 10 | [MiniMax M3.1 Flash](https://benchize.com/models/minimax-m3-1-flash-max.md)Max | 25.25% | SE ±1.01 pp | $0.73 | 3,030.55source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 25.253 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.734722 \| USD \| per_test \| Lower is better \| \| `latency` \| 3030.551 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort max harness Not reported maxOutputTokens 262144 provider MiniMax reasoning Not reported Reasoning effort max Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.01 pp sourcePopulation 23 stderr 1.01 tokenTotals {"cache_read_tokens":2489516469,"cache_write_tokens":0,"input_tokens":2497428556,"output_tokens":16603505,"reasoning_tokens":13158836} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label minimax/MiniMax-M3.1-flash Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/minimax~1MiniMax-M3.1-flash` | | | | | | | | 11 | [MiMo-V2.6-Flash](https://benchize.com/models/mimo-v2-6-flash.md) | 24.24% | SE ±1.52 pp | $0.22 | 8,712.24source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 24.242 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.216275 \| USD \| per_test \| Lower is better \| \| `latency` \| 8712.24 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider Xiaomi reasoning Not reported Reasoning effort unknown Temperature 1 topP 0.95 verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.52 pp sourcePopulation 23 stderr 1.515 tokenTotals {"cache_read_tokens":2150748075,"cache_write_tokens":0,"input_tokens":2175333414,"output_tokens":16125095,"reasoning_tokens":12753311} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label xiaomi/mimo-v2.6-flash Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/xiaomi~1mimo-v2.6-flash` | | | | | | | | 12 | [Gemini 3.8 Flash](https://benchize.com/models/gemini-3-8-flash-high.md)High | 19.7% | SE ±2.31 pp | $8.65 | 6,427.58source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 19.697 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 8.64749 \| USD \| per_test \| Lower is better \| \| `latency` \| 6427.575 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 65536 provider Google reasoning Not reported Reasoning effort high Temperature 1 topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.31 pp sourcePopulation 23 stderr 2.314 tokenTotals {"cache_read_tokens":2313842969,"cache_write_tokens":0,"input_tokens":2400090539,"output_tokens":12571537,"reasoning_tokens":7741941} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label google/gemini-3.8-flash Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/google~1gemini-3.8-flash` | | | | | | | | 13 | [DeepSeek V4 Pro 0813](https://benchize.com/models/deepseek-v4-pro-0813-max.md)Max | 14.14% | SE ±2.02 pp | $3.31 | 5,493.6source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 14.141 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 3.311817 \| USD \| per_test \| Lower is better \| \| `latency` \| 5493.603 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 384000 provider DeepSeek reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±2.02 pp sourcePopulation 23 stderr 2.02 tokenTotals {"cache_read_tokens":3587685589,"cache_write_tokens":0,"input_tokens":3598466519,"output_tokens":11740127,"reasoning_tokens":6902591} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label deepseek/deepseek-v4-pro-0813 Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/deepseek~1deepseek-v4-pro-0813` | | | | | | | | 14 | [GPT-6 Luna](https://benchize.com/models/gpt-6-luna-max.md)Max | 13.64% | SE ±1.52 pp | $0.35 | 1,965.57source duration | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 13.636 \| percent \| reported \| Higher is better \| \| `cost_per_test` \| 0.354726 \| USD \| per_test \| Lower is better \| \| `latency` \| 1965.568 \| source_duration \| reported_duration \| Lower is better \| ### Configuration computeEffort Not reported harness Not reported maxOutputTokens 128000 provider OpenAI reasoning Not reported Reasoning effort max Temperature Not reported topP Not reported verbosity Not reported ### Reported statistics Uncertainty Standard error: ±1.52 pp sourcePopulation 23 stderr 1.515 tokenTotals {"cache_read_tokens":988222853,"cache_write_tokens":15670624,"input_tokens":1004162895,"output_tokens":10957435,"reasoning_tokens":7328334} ### Source [Original source](https://www.vals.ai/_astro/benchmark_view_vals_index.ye6C1wH2.json) Source Vals AI Original model label openai/gpt-6-luna Source version 2.2 Observed 2026-10-21T16:45:49.430Z Evidence SHA-256 `0e95aee94cd2bf34bc4b43a8f4bb16c7817de138613c21011873fea4e0786b45` Source locator `/tasks/terminal_bench_4/openai~1gpt-6-luna` | | | | | | | --- Cite as: Vals Index / terminal bench 4 — Benchize, https://benchize.com/benchmarks/vals-ai%3Avals_index%3Aterminal_bench_4 (retrieved 2026-10-21)