MODEL COMPARISON

Find strengths. Understand tradeoffs.

Compare up to four exact model configurations across the full benchmark collection, with the evidence beside every score.

View radar ↓Full benchmark table ↓

Composite retains Coding Agent v1.4 · source 2026-09-09

The five Composite inputs remain unchanged. Its Coding Agent input is the median across complete harness results in the retained v1.4 snapshot from 2026-09-09. Artificial Analysis now publishes v1.5, with different components. Explore current v1.5 separately. Dated snapshot labels on other indices identify unversioned source captures, not a verified semantic version.

BUILD YOUR COMPARISON

Choose up to four models

2 / 4 selected

Exact reasoning configurations. All catalog models are available; price and provider filters do not hide benchmark evidence.

Find models

Model A

Benchmark sheet ↗

Model B

Benchmark sheet ↗

Model C

Model D

2 models selected. 23 evaluation rows in the full comparison.

Radar axes · 6 / 8 selected

Choose 3–8 axes. Every option names one benchmark version and evaluation group. Unmeasured selections remain empty.

RELATIVE PERFORMANCE

Benchmark radar

Independent measurements

Each version uses its measured catalog minimum → 0 and maximum → 100; lower-is-better axes are reversed. Missing scores leave gaps. A zero is a measured minimum, not missing.

A · GPT-5.6 Sol (high)
B · Claude Sonnet 5 (Adaptive Reasoning, High Effort)

Scroll the chart horizontally. All values are also in the table below.

Radar values · normalized / 100, rounded to 3 decimals (exact native score). Missing or uninformative axes are never filled.
Axis / version	A · GPT-5.6 Sol (high)	B · Claude Sonnet 5 (Adaptive Reasoning, High Effort)	Normalization range
1. GPQA Diamond (AA) Version snapshot-2026-09-10 · Published board	96.028 (0.928282828282828 fraction)	No measured result	612 measured configurations · 0.097979797979798–0.962626262626263 fraction · higher better
2. Humanity's Last Exam (AA text-only) Version snapshot-2026-09-10 · Published board	77.414 (0.46014828544949 fraction)	59.697 (0.357275254865616 fraction)	607 measured configurations · 0.0106580166821131–0.591288229842447 fraction · higher better
3. AA-LCR v1.1 Version 1.1 · Published board	92.105 (0.816666666666667 fraction)	86.466 (0.766666666666667 fraction)	517 measured configurations · 0–0.886666666666667 fraction · higher better
4. SciCode (AA subproblems) Version 1.0.1 · Published board	89.074 (0.577546296296296 fraction)	81.948 (0.542824074074074 fraction)	167 measured configurations · 0.143518518518519–0.630787037037037 fraction · higher better
5. AA Coding Index Version snapshot-2026-09-10 (unversioned) · Published board	94.608 (77.2 points)	No measured result	256 measured configurations · 0–81.6 points · higher better
6. AA Intelligence Index Version snapshot-2026-09-10 (unversioned) · Published board	78.024 (42.5 points)	No measured result	633 measured configurations · 3.8–53.4 points · higher better

How to read this chart

Formula: 100 × (value − minimum) / (maximum − minimum), or 100 minus that value for lower-is-better axes. Peers are independently measured, exactly matched catalog configurations; repeated source identities count once. Fewer than two peers, a constant range or unknown direction suppresses plotting. DesignArena results below 200 battles are excluded. Axes show version and published harness/configuration. Model prompts and test conditions can still differ; inspect the source before interpreting small differences. Ranges stay fixed when model selection changes. This chart does not change Composite.

EVERY COLLECTED BENCHMARK

Full benchmark comparison

Native units; versions and evaluation groups remain separate. Measured results take priority over vendor claims, then the latest observation. Small differences are not evidence of significance.

Find a benchmarkCategoryInclude rows without selected-model results

All matching benchmark versions and model scores with evidence
Benchmark / version	A · GPT-5.6 Sol (high)	B · Claude Sonnet 5 (Adaptive Reasoning, High Effort)
Agentic
AA-Briefcase Version snapshot-2026-09-10 Published board Elo · higher better	1360.88 Elomeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-Briefcase · snapshot-2026-09-10 · Published board Exact value: `1360.88` Elo Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:briefcaseBreakdown.overall.elo` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	1177.84 Elomeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-Briefcase · snapshot-2026-09-10 · Published board Exact value: `1177.84` Elo Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:briefcaseBreakdown.overall.elo` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
GDPval-AA v2 Version 2 Published board Elo · higher better	1524.01 Elomeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: GDPval-AA v2 · 2 · Published board Exact value: `1524.01` Elo Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:gdpval` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	1308.45 Elomeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: GDPval-AA v2 · 2 · Published board Exact value: `1308.45` Elo Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:gdpval` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
Terminal-Bench Hard (AA) Version 74221fb Published board fraction · higher better	0.62121 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Terminal-Bench Hard (AA) · 74221fb · Published board Exact value: `0.621212121212121` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:terminalbenchHard` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Terminal-Bench v2.1 (AA) Version 2.1 Published board fraction · higher better	0.87266 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Terminal-Bench v2.1 (AA) · 2.1 · Published board Exact value: `0.872659176029963` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:terminalbenchV21` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Terminal-Bench v4.0 (AA) Version 4.0 Published board fraction · higher better	0.20707 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Terminal-Bench v4.0 (AA) · 4.0 · Published board Exact value: `0.207070707070707` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:terminalbenchV40` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.05051 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Terminal-Bench v4.0 (AA) · 4.0 · Published board Exact value: `0.0505050505050505` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:terminalbenchV40` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
Coding
Artificial Analysis Coding Agent Index v1.4 Version 1.4 Codex fraction · higher better	0.64116 fractionmeasuredobserved 2026-09-09artificialanalysis.ai ↗ Evidence Axis: Artificial Analysis Coding Agent Index v1.4 · 1.4 · Codex Exact value: `0.6411644486641197` fraction Observed: 2026-09-09 · publication date: not recorded Observation id: `aa-coding:1.4:35` Source: https://artificialanalysis.ai/ Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
SciCode (AA subproblems) Version 1.0.1 Published board fraction · higher better	0.57755 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: SciCode (AA subproblems) · 1.0.1 · Published board Exact value: `0.577546296296296` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:scicode` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.54282 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: SciCode (AA subproblems) · 1.0.1 · Published board Exact value: `0.542824074074074` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:scicode` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
AA Coding Index Version snapshot-2026-09-10 (unversioned) Published board points · higher better	77.2 pointsmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA Coding Index · snapshot-2026-09-10 (unversioned) · Published board Exact value: `77.2` points Observed: 2026-09-10 · publication date: not recorded Observation id: `legacy:aa_coding_index:gpt-5.6-sol::high` Source: https://artificialanalysis.ai/leaderboards/models Protocol: inspect · file `data/raw/artificialanalysis.json`	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
DesignArena Frontend Version snapshot-2026-09-10 (unversioned) Published board Elo · higher better	1209 Elomeasured2255 battlesobserved 2026-09-10designarena.ai ↗ Evidence Axis: DesignArena Frontend · snapshot-2026-09-10 (unversioned) · Published board Exact value: `1209` Elo Observed: 2026-09-10 · publication date: not recorded Observation id: `legacy:frontend:gpt-5.6-sol::high` Source: https://www.designarena.ai/ Protocol: inspect · file `data/raw/designarena.json`	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
DesignArena Full-Stack Version snapshot-2026-09-10 (unversioned) Published board Elo · higher better	1168 Elomeasured1611 battlesobserved 2026-09-10designarena.ai ↗ Evidence Axis: DesignArena Full-Stack · snapshot-2026-09-10 (unversioned) · Published board Exact value: `1168` Elo Observed: 2026-09-10 · publication date: not recorded Observation id: `legacy:fullstack:gpt-5.6-sol::high` Source: https://www.designarena.ai/ Protocol: inspect · file `data/raw/designarena.json`	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Instruction-following
IFBench (AA single-turn) Version snapshot-2026-09-10 Published board fraction · higher better	0.69184 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: IFBench (AA single-turn) · snapshot-2026-09-10 · Published board Exact value: `0.691836734693878` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:ifbench` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Knowledge
Humanity's Last Exam (AA text-only) Version snapshot-2026-09-10 Published board fraction · higher better	0.46015 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Humanity's Last Exam (AA text-only) · snapshot-2026-09-10 · Published board Exact value: `0.46014828544949` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:hle` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.35728 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: Humanity's Last Exam (AA text-only) · snapshot-2026-09-10 · Published board Exact value: `0.357275254865616` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:hle` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
AA-Omniscience Index Version snapshot-2026-09-10 Published board points · higher better	20.36667 pointsmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-Omniscience Index · snapshot-2026-09-10 · Published board Exact value: `20.3666666666667` points Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:omniscience` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	-3.68333 pointsmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-Omniscience Index · snapshot-2026-09-10 · Published board Exact value: `-3.68333333333333` points Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:omniscience` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
Long-context
GDP.pdf (AA) Version snapshot-2026-09-10 Published board fraction · higher better	0.278 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: GDP.pdf (AA) · snapshot-2026-09-10 · Published board Exact value: `0.278` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:gdpPdfAllPass` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.094 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: GDP.pdf (AA) · snapshot-2026-09-10 · Published board Exact value: `0.094` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:gdpPdfAllPass` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
AA-LCR v1.1 Version 1.1 Published board fraction · higher better	0.81667 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-LCR v1.1 · 1.1 · Published board Exact value: `0.816666666666667` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:lcr` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.76667 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA-LCR v1.1 · 1.1 · Published board Exact value: `0.766666666666667` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:lcr` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
MLCR-AA Version snapshot-2026-09-10 Published board fraction · higher better	0.17222 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: MLCR-AA · snapshot-2026-09-10 · Published board Exact value: `0.172222222222222` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:mlcrOverall` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Reasoning
AA Intelligence Index Version snapshot-2026-09-10 (unversioned) Published board points · higher better	42.5 pointsmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AA Intelligence Index · snapshot-2026-09-10 (unversioned) · Published board Exact value: `42.5` points Observed: 2026-09-10 · publication date: not recorded Observation id: `legacy:aa_intelligence_index:gpt-5.6-sol::high` Source: https://artificialanalysis.ai/leaderboards/models Protocol: inspect · file `data/raw/artificialanalysis.json`	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Science
CritPt (AA) Version snapshot-2026-09-10 Published board fraction · higher better	0.25714 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: CritPt (AA) · snapshot-2026-09-10 · Published board Exact value: `0.257142857142857` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:critpt` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.15143 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: CritPt (AA) · snapshot-2026-09-10 · Published board Exact value: `0.151428571428571` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:critpt` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
GPQA Diamond (AA) Version snapshot-2026-09-10 Published board fraction · higher better	0.92828 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: GPQA Diamond (AA) · snapshot-2026-09-10 · Published board Exact value: `0.928282828282828` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:gpqa` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Tool-use
AutomationBench-AA Version 1.0.6 Published board fraction · higher better	0.55312 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AutomationBench-AA · 1.0.6 · Published board Exact value: `0.5531220357393729` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:automationBenchPartialScore` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	0.32097 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: AutomationBench-AA · 1.0.6 · Published board Exact value: `0.32096638814380396` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:ba0224cf-0351-4f56-8508-b3f1a740ae4a:automationBenchPartialScore` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗
τ²-Bench Telecom (AA) Version snapshot-2026-09-10 Published board fraction · higher better	0.83333 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: τ²-Bench Telecom (AA) · snapshot-2026-09-10 · Published board Exact value: `0.833333333333333` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:tau2` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
τ³-Banking (AA) Version 1.0.1 Published board fraction · higher better	0.36701 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: τ³-Banking (AA) · 1.0.1 · Published board Exact value: `0.36701030927835` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:tauBanking` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.
Vision
MMMU Pro (AA) Version snapshot-2026-09-10 Published board fraction · higher better	0.8185 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗ Evidence Axis: MMMU Pro (AA) · snapshot-2026-09-10 · Published board Exact value: `0.81849710982659` fraction Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded Observation id: `aa:8afc250d-b538-45a2-812a-4605f4ffd87e:mmmuPro` Source: https://artificialanalysis.ai/models/gpt-5-6-sol Exact observation and full protocol ↗	Unknown / no published match No observation is attached to this exact catalog configuration. This does not establish that the model was never tested.

GPT-5.6 Sol (high)

Profile signals

1 threshold-crossing signal flagged · 21 eligible benchmark families

How flags are calculated

Heuristic screen, not statistical significance: benchmark families are correlated and source uncertainty is unknown. Peer evidence needs ≥ 20 independently measured matched configurations from ≥ 10 distinct model families. A flag needs a directed population z-score of magnitude ≥ 1.5 and a gap of ≥ 1.5 from the leave-one-benchmark-family-out mean z in the same direction, over ≥ 5 other benchmark families.

unusually strong CritPt (AA) snapshot-2026-09-10
Why
Observed: 0.25714 fraction
Peer mean 0.03942 · peer sd 0.0762 · n 521 · families 367
Directed z 2.857 · baseline z 1.225 · gap 1.632 · profile n 20
0.25714 fractionmeasuredobserved 2026-09-10artificialanalysis.ai ↗
Evidence
Axis: CritPt (AA) · snapshot-2026-09-10 · Published board
Exact value: 0.257142857142857 fraction
Observed: 2026-09-10T21:47:16.627Z · publication date: not recorded
Observation id: aa:8afc250d-b538-45a2-812a-4605f4ffd87e:critpt
Source: https://artificialanalysis.ai/models/gpt-5-6-sol
Exact observation and full protocol ↗

Protocol-compatible measured/vendor divergences

No verified protocol-compatible vendor/measured pair is available for this model; agreement cannot be assessed.

Claude Sonnet 5 (Adaptive Reasoning, High Effort)

Profile signals

No threshold-crossing signal · 10 eligible benchmark families

How flags are calculated

Protocol-compatible measured/vendor divergences

No verified protocol-compatible vendor/measured pair is available for this model; agreement cannot be assessed.

Price & provider comparison · separate two-model selection

This price view has its own A/B selection and uses the global price and provider filters. Benchmark exploration above includes the full catalog.

Loading price and provider evidence…