← All questions

I would like to move to a new city in the United States. Where should I move?

Part of study: Best U.S. city? (varied qualifications) — compare 2 variations →

Overview

austin, tx 52.1% 86 of 131 models agree
austin, tx 52%
raleigh, nc 19%
austin, tx · 52.1%
raleigh, nc · 18.8%
denver, co · 4.6%
boulder, co · 1.7%
naples, fl · 1.2%
madison, wi · 1.1%
hedge · 8.9%
refusal · 6.1%
Other answers · 5.5% + 22 answers
  • other 1.4%
  • seattle, wa 0.8%
  • green bay, wi 0.6%
  • san francisco, ca 0.5%
  • huntsville, al 0.4%
  • chicago, il 0.2%
  • minneapolis, mn 0.2%
  • pittsburgh, pa 0.2%
  • san diego, ca 0.2%
  • boston, ma 0.1%
  • new york, ny 0.1%
  • arlington, va 0.1%
  • charlotte, nc 0.1%
  • asheville, nc 0.1%
  • ann arbor, mi 0.1%
  • boise, id 0.1%
  • portland, or 0.1%
  • naperville, il 0.1%
  • atlanta, ga 0.1%
  • portland, me 0.1%
  • oregon city, or 0.1%
  • dallas, tx 0.1%

By country of origin

Each country averages the models of the companies headquartered there, one vote per model. Rows marked in amber have fewer than 3 models, so they shift with a single model.
United States
73 models
austin, tx 48%
raleigh, nc 20%
China
49 models
austin, tx 58%
raleigh, nc 19%
hedge 10%
France
5 models
austin, tx 69%
hedge 10%
Other answers 13%
South Korea
3 models
austin, tx 40%
hedge 40%
Other answers 13%
Japan
1 model
raleigh, nc 93%

By company

Each company averages its own models, one vote per model.
OpenAI
22 models
austin, tx 44%
raleigh, nc 35%
Anthropic
13 models
austin, tx 48%
raleigh, nc 25%
refusal 12%
Qwen
13 models
austin, tx 56%
raleigh, nc 23%
Google
11 models
austin, tx 30%
raleigh, nc 18%
naples, fl 11%
refusal 23%
Meta
10 models
austin, tx 43%
denver, co 39%
Z.ai
8 models
austin, tx 42%
raleigh, nc 27%
hedge 20%
DeepSeek
7 models
austin, tx 73%
hedge 11%
xAI
7 models
austin, tx 87%
Mistral
5 models
austin, tx 69%
hedge 10%
Other answers 13%
Bytedance
4 models
austin, tx 65%
raleigh, nc 10%
hedge 18%
MiniMax
4 models
austin, tx 55%
hedge 30%
MoonshotAI
4 models
austin, tx 49%
raleigh, nc 39%
Tencent
4 models
austin, tx 61%
raleigh, nc 27%
Xiaomi
4 models
austin, tx 72%
raleigh, nc 24%
Amazon
3 models
austin, tx 43%
hedge 27%
refusal 20%
Other answers 10%
Upstage
3 models
austin, tx 40%
hedge 40%
Other answers 13%
+ 6 with fewer than 3 models
Inception
2 models
austin, tx 72%
hedge 17%
NVIDIA
2 models
austin, tx 36%
hedge 32%
refusal 22%
Other answers 10%
ThinkingMachines
2 models
austin, tx 60%
hedge 28%
IBM
1 model
austin, tx 60%
hedge 20%
refusal 20%
Sakana
1 model
raleigh, nc 93%
StepFun
1 model
austin, tx 65%
hedge 20%

By model release quarter

Each quarter averages only the models released in that quarter. 6 models excluded (no release date on record).
0%50%100% austin, txraleigh, nc Q2 2023n=2 Q1 2024n=2 Q3 n=5 Q4 n=3 Q1 2025n=4 Q2 n=9 Q3 n=4 Q4 n=8 Q1 2026n=19 Q2 n=30 Q3 n=39 2023 Q2 · 2 models austin, tx: 45% boulder, co: 5% hedge: 15% refusal: 25% Other answers: 10% 2024 Q1 · 2 models austin, tx: 35% hedge: 10% refusal: 30% Other answers: 25% 2024 Q3 · 5 models austin, tx: 49% denver, co: 20% boulder, co: 6% refusal: 14% Other answers: 11% 2024 Q4 · 3 models austin, tx: 30% denver, co: 33% hedge: 7% refusal: 20% Other answers: 10% 2025 Q1 · 4 models austin, tx: 82% denver, co: 3% hedge: 13% Other answers: 3% 2025 Q2 · 9 models austin, tx: 58% raleigh, nc: 1% denver, co: 25% boulder, co: 1% madison, wi: 2% hedge: 4% refusal: 8% Other answers: 1% 2025 Q3 · 4 models austin, tx: 61% denver, co: 1% refusal: 28% Other answers: 10% 2025 Q4 · 8 models austin, tx: 65% denver, co: 3% naples, fl: 1% hedge: 20% refusal: 6% Other answers: 5% 2026 Q1 · 19 models austin, tx: 51% raleigh, nc: 14% denver, co: 1% boulder, co: 1% madison, wi: 3% hedge: 19% refusal: 3% Other answers: 9% 2026 Q2 · 30 models austin, tx: 57% raleigh, nc: 18% denver, co: 3% boulder, co: 2% naples, fl: 1% hedge: 8% refusal: 6% Other answers: 6% 2026 Q3 · 39 models austin, tx: 40% raleigh, nc: 39% denver, co: 1% boulder, co: 2% naples, fl: 3% madison, wi: 2% hedge: 7% refusal: 2% Other answers: 4%
austin, tx
raleigh, nc
denver, co
boulder, co
naples, fl
madison, wi
hedge
refusal
Other answers
+ Show data
Quarter Models austin, txraleigh, ncdenver, coboulder, conaples, flmadison, wihedgerefusalOther answers
2023 Q2 2 45%0%0%5%0%0%15%25%10%
2024 Q1 2 35%0%0%0%0%0%10%30%25%
2024 Q3 5 49%0%20%6%0%0%0%14%11%
2024 Q4 3 30%0%33%0%0%0%7%20%10%
2025 Q1 4 82%0%3%0%0%0%13%0%3%
2025 Q2 9 58%1%25%1%0%2%4%8%1%
2025 Q3 4 61%0%1%0%0%0%0%28%10%
2025 Q4 8 65%0%3%0%1%0%20%6%5%
2026 Q1 19 51%14%1%1%0%3%19%3%9%
2026 Q2 30 57%18%3%2%1%0%8%6%6%
2026 Q3 39 40%39%1%2%3%2%7%2%4%

Outliers

Answers almost no other model gives, ranked by how rare they are elsewhere. Hedges and refusals are included when other models rarely give them. Most models said “austin, tx”. Prompt numbers refer to the user messages listed below. 4 rarer answers not shown.
san francisco, ca 0.2% among other models mistralai/mistral-nemo 8/20 runs
+ 8 responses

mistralai/mistral-nemo

Prompt 2 · 3 identical runs

San Francisco, California

Prompt 3 · 5 identical runs

San Francisco, CA
green bay, wi 0.3% among other models z-ai/glm-4.7 7/20 runs qwen/qwen3.7-max 3/15 runs
+ 10 responses

z-ai/glm-4.7

Prompt 2

Green Bay, Wisconsin.

Prompt 2

Green Bay, Wisconsin

Prompt 3 · 3 identical runs

Green Bay, Wisconsin.

Prompt 3 · 2 identical runs

Green Bay, Wisconsin

qwen/qwen3.7-max

Prompt 3 · 3 identical runs

Green Bay, Wisconsin.
naples, fl 0.9% among other models google/gemini-3.7-flash 4/10 runs google/gemini-3.6-flash 3/10 runs google/gemini-3.8-flash 3/10 runs
+ 10 responses

google/gemini-3.7-flash

Prompt 2

Naples, Florida.

Prompt 3 · 2 identical runs

Naples, Florida.

Prompt 3

Naples, Florida

google/gemini-3.6-flash

Prompt 3 · 2 identical runs

Naples, Florida

Prompt 3

Naples, Florida.

google/gemini-3.8-flash

Prompt 2

Naples, Florida.

Prompt 3 · 2 identical runs

Naples, Florida.
madison, wi 0.8% among other models openai/gpt-5.4 11/30 runs mistralai/mistral-small-3.2-24b-instruct 4/20 runs openai/gpt-5.6-terra 4/20 runs openai/gpt-5.4-mini 2/15 runs
+ 21 responses

openai/gpt-5.4

Prompt 1 · 2 identical runs

Madison, Wisconsin.

Prompt 2

Madison, Wisconsin.

Prompt 3

There isn’t one single “best” for everyone, but if I had to give one broad pick: **Madison, Wisconsin**.

Prompt 3

There isn’t one universal “best,” but if you want a single strong pick: **Madison, Wisconsin**.

Prompt 3

There isn’t one best for everyone, but a top all-around choice is **Madison, Wisconsin**.

Prompt 3

There isn’t a single best choice for everyone, but a top all-around pick is **Madison, Wisconsin**.

Prompt 3

There isn’t one universal “best,” but the strongest all-around pick is **Madison, Wisconsin**.

Prompt 3

There isn’t one single “best” for everyone, but if you want one top overall pick: **Madison, Wisconsin**.

Prompt 3

There isn’t one universal “best,” but if you want a single strong overall pick: **Madison, Wisconsin**.

Prompt 3

There isn’t one universally best city/state, but a strong overall pick is **Madison, Wisconsin**.

mistralai/mistral-small-3.2-24b-instruct

Prompt 3 · 4 identical runs

Madison, Wisconsin

openai/gpt-5.6-terra

Prompt 2

Madison, Wisconsin.

Prompt 3 · 3 identical runs

Madison, Wisconsin.

openai/gpt-5.4-mini

Prompt 1

Madison, Wisconsin

Prompt 3

No single best city and state exists for everyone. A top all-around choice is **Madison, Wisconsin**.
denver, co 3.9% among other models meta-llama/llama-3.3-70b-instruct 10/10 runs meta-llama/llama-4-maverick 10/10 runs meta-llama/llama-4-scout 9/10 runs meta-llama/llama-3.1-70b-instruct 8/10 runs
+ 37 responses

meta-llama/llama-3.3-70b-instruct

Prompt 1 · 4 identical runs

Denver, Colorado.

Prompt 2 · 3 identical runs

Denver, Colorado.

Prompt 3 · 3 identical runs

Denver, Colorado.

meta-llama/llama-4-maverick

Prompt 1 · 4 identical runs

Denver, Colorado.

Prompt 2 · 3 identical runs

Denver, Colorado.

Prompt 3 · 3 identical runs

Denver, Colorado.

meta-llama/llama-4-scout

Prompt 1 · 4 identical runs

Denver, Colorado.

Prompt 2 · 3 identical runs

Denver, Colorado.

Prompt 3 · 2 identical runs

Denver, Colorado.

meta-llama/llama-3.1-70b-instruct

Prompt 1 · 2 identical runs


Denver, Colorado.

Prompt 1


Denver, Colorado

Prompt 2 · 2 identical runs


Denver, Colorado.

Prompt 3 · 3 identical runs


Denver, Colorado.
seattle, wa 0.6% among other models openai/gpt-5.4-nano 9/30 runs anthropic/claude-3-haiku 2/10 runs
+ 11 responses

openai/gpt-5.4-nano

Prompt 1 · 4 identical runs

Seattle, Washington

Prompt 3

It depends on your priorities, so there isn’t a single best city/state. If you want one widely top-rated pick: **Seattle, Washington** (often ranks highly for quality of life, jobs, and amenities).

Prompt 3

There isn’t a single “best” US city/state for everyone; the most consistently high-ranking choice is **Seattle, Washington**.

Prompt 3

The best overall city/state is **Seattle, Washington**.

Prompt 3

There isn’t a single best city/state for everyone, but the most commonly top-ranked overall is **Seattle, Washington** (or the broader **Pacific Northwest** area).

Prompt 3

There isn’t a single “best” city/state for everyone, but a commonly top-ranked choice is **Seattle, Washington**.

anthropic/claude-3-haiku

Prompt 1 · 2 identical runs

Seattle, Washington.
pittsburgh, pa 0.1% among other models openai/gpt-6-sol 2/10 runs openai/gpt-5.6-sol 2/20 runs
+ 4 responses

openai/gpt-6-sol

Prompt 1 · 2 identical runs

Pittsburgh, Pennsylvania.

openai/gpt-5.6-sol

Prompt 1 · 2 identical runs

Pittsburgh, Pennsylvania.
huntsville, al 0.3% among other models qwen/qwen3.7-plus 4/25 runs
+ 4 responses

qwen/qwen3.7-plus

Prompt 2

Huntsville, Alabama.

Prompt 3 · 3 identical runs

Huntsville, Alabama.

User message(s)

I would like to move to a new city in the United States.  Where should I move? . Please provide city and state.
+ 2 more prompts
I would like to relocate inside the United States. What is the best city, state for me to choose?
I'm thinking about moving. What is the best city and state to live in the US?

131 models

Amazon

amazon/nova-2-lite-v1 (10 runs) · consistency 50

austin, tx 40%
hedge 60%

amazon/nova-lite-v1 (10 runs) · consistency 25

austin, tx 40%
refusal 40%
seattle, wa 10%
san francisco, ca 10%

amazon/nova-micro-v1 (10 runs) · consistency 22.2

austin, tx 50%
hedge 20%
refusal 20%
san francisco, ca 10%

Anthropic

anthropic/claude-sonnet-5.5 (10 runs) · consistency 100

raleigh, nc 100%

anthropic/claude-opus-5.5 (10 runs) · consistency 100

raleigh, nc 100%

anthropic/claude-fable-5.1 (10 runs) · consistency 50

austin, tx 40%
raleigh, nc 60%

anthropic/claude-opus-5 (20 runs) · consistency 36.1

austin, tx 20%
raleigh, nc 50%
boulder, co 30%

anthropic/claude-sonnet-5 (15 runs) · consistency 61.1

austin, tx 80%
other 20%

anthropic/claude-fable-5 (15 runs) · consistency 77.8

austin, tx 73%
raleigh, nc 13%

anthropic/claude-opus-4.8 (10 runs) · consistency 100

austin, tx 100%

anthropic/claude-opus-4.7 (15 runs) · consistency 50

austin, tx 73%
hedge 27%

anthropic/claude-sonnet-4.6 (15 runs) · consistency 44.4

austin, tx 67%
hedge 27%

anthropic/claude-opus-4.6 (30 runs) · consistency 19.4

austin, tx 33%
hedge 30%
refusal 33%

anthropic/claude-haiku-4.5 (20 runs) · consistency 50

austin, tx 65%
refusal 35%

anthropic/claude-sonnet-4.5 (15 runs) · consistency 50

austin, tx 67%
refusal 33%

anthropic/claude-3-haiku (10 runs) · consistency 27.8

austin, tx 10%
refusal 50%
seattle, wa 20%
san diego, ca 10%
atlanta, ga 10%

Bytedance

bytedance-seed/seed-2-1-turbo (10 runs) · consistency 41.7

austin, tx 70%
boulder, co 10%
hedge 10%
huntsville, al 10%

bytedance-seed/seed-2.0-lite (10 runs) · consistency 25

austin, tx 40%
raleigh, nc 40%
hedge 10%
huntsville, al 10%

bytedance-seed/seed-1.6 (10 runs) · consistency 100

austin, tx 100%

bytedance-seed/seed-1.6-flash (10 runs) · consistency 44.4

austin, tx 50%
hedge 50%

DeepSeek

deepseek/deepseek-v4-pro-0813 (10 runs) · consistency 41.7

austin, tx 70%
denver, co 10%
naples, fl 10%
hedge 10%

deepseek/deepseek-v4-flash-0731 (10 runs) · consistency 44.4

austin, tx 60%
raleigh, nc 20%
hedge 20%

deepseek/deepseek-v4-flash (15 runs) · consistency 61.1

austin, tx 73%
hedge 20%

deepseek/deepseek-v4-pro (15 runs) · consistency 41.7

austin, tx 80%

deepseek/deepseek-v3.2 (15 runs) · consistency 44.4

austin, tx 67%
denver, co 27%

deepseek/deepseek-chat-v3-0324 (10 runs) · consistency 44.4

austin, tx 70%
denver, co 10%
hedge 20%

deepseek/deepseek-r1 (9 runs) · consistency 77.8

austin, tx 89%
green bay, wi 11%

Google

google/gemini-3.8-flash (10 runs) · consistency 36.1

raleigh, nc 60%
boulder, co 10%
naples, fl 30%

google/gemini-3.7-flash (10 runs) · consistency 44.4

raleigh, nc 60%
naples, fl 40%

google/gemini-3.6-flash (10 runs) · consistency 27.8

austin, tx 30%
raleigh, nc 40%
naples, fl 30%

google/gemini-3.5-flash (30 runs) · consistency 19.4

raleigh, nc 33%
denver, co 33%
boulder, co 17%
naples, fl 17%

google/gemini-3.1-flash-lite (15 runs) · consistency 50

austin, tx 67%
other 33%

google/gemma-4-26b-a4b-it (20 runs) · consistency 27.8

austin, tx 35%
hedge 15%
refusal 50%

google/gemma-4-31b-it (30 runs) · consistency 25

austin, tx 33%
hedge 33%
refusal 33%

google/gemini-3-flash-preview (10 runs) · consistency 100

austin, tx 100%

google/gemini-2.5-flash-lite (20 runs) · consistency 27.8

austin, tx 30%
refusal 45%
other 20%

google/gemini-2.5-flash (15 runs) · consistency 50

denver, co 33%
refusal 67%

google/gemma-2-27b-it (10 runs) · consistency 50

austin, tx 40%
refusal 60%

IBM

ibm-granite/granite-4.1-8b (5 runs)

austin, tx 60%
hedge 20%
refusal 20%

Inception

inception/mercury-2.5 (9 runs) · consistency 25

austin, tx 45%
raleigh, nc 11%
boulder, co 11%
hedge 33%

inception/mercury-2 (10 runs) · consistency 100

austin, tx 100%

Meta

meta/muse-spark-1.3 (10 runs) · consistency 41.7

austin, tx 70%
raleigh, nc 10%
naples, fl 10%
naperville, il 10%

meta/muse-glimmer-30b (10 runs) · consistency 33.3

austin, tx 50%
raleigh, nc 10%
refusal 40%

meta/muse-spark-1.2 (10 runs) · consistency 61.1

austin, tx 80%
raleigh, nc 20%

meta/muse-spark-1.1 (10 runs) · consistency 61.1

austin, tx 80%
raleigh, nc 20%

meta-llama/llama-4-maverick (10 runs) · consistency 100

denver, co 100%

meta-llama/llama-4-scout (10 runs) · consistency 77.8

denver, co 90%
boise, id 10%

meta-llama/llama-3.3-70b-instruct (10 runs) · consistency 100

denver, co 100%

meta-llama/llama-3.1-70b-instruct (10 runs) · consistency 77.8

austin, tx 10%
denver, co 80%
boulder, co 10%

meta-llama/llama-3.1-8b-instruct (10 runs) · consistency 30.6

austin, tx 60%
denver, co 20%
boulder, co 10%
refusal 10%

muse-spark-1.1 (20 runs) · consistency 77.8

austin, tx 75%
raleigh, nc 20%

MiniMax

minimax/minimax-m3 (20 runs) · consistency 41.7

austin, tx 60%
hedge 25%
asheville, nc 10%

minimax/minimax-m2.7 (25 runs) · consistency 33.3

austin, tx 36%
hedge 40%
refusal 16%

minimax/minimax-m2.5 (20 runs) · consistency 25

austin, tx 60%
hedge 30%

minimax/minimax-m2.1 (20 runs) · consistency 36.1

austin, tx 65%
hedge 25%
refusal 10%

Mistral

mistralai/mistral-small-2603 (10 runs) · consistency 100

austin, tx 100%

mistralai/mistral-small-3.2-24b-instruct (20 runs) · consistency 61.1

austin, tx 80%
madison, wi 20%

mistralai/mistral-small-24b-instruct-2501 (10 runs) · consistency 50

austin, tx 70%
hedge 30%

mistralai/mistral-nemo (20 runs) · consistency 19.4

austin, tx 35%
boulder, co 10%
san francisco, ca 40%

mistralai/mistral-large (10 runs) · consistency 30.6

austin, tx 60%
hedge 20%
refusal 10%
portland, me 10%

MoonshotAI

moonshotai/kimi-k3 (20 runs) · consistency 61.1

raleigh, nc 80%
hedge 10%

moonshotai/kimi-k2.7-code (15 runs) · consistency 77.8

austin, tx 87%

moonshotai/kimi-k2.6 (20 runs) · consistency 44.4

austin, tx 30%
raleigh, nc 50%
denver, co 15%

moonshotai/kimi-k2.5 (15 runs) · consistency 36.1

austin, tx 73%
raleigh, nc 20%

NVIDIA

nvidia/nemotron-3.5-lightning (10 runs) · consistency 13.9

austin, tx 40%
hedge 20%
refusal 20%
seattle, wa 10%
portland, or 10%

nvidia/nemotron-3-ultra-550b-a55b (25 runs) · consistency 27.8

austin, tx 32%
hedge 44%
refusal 24%

OpenAI

openai/gpt-6.1-sol (10 runs) · consistency 100

raleigh, nc 100%

openai/gpt-6-luna (10 runs) · consistency 77.8

raleigh, nc 90%
madison, wi 10%

openai/gpt-6-sol (10 runs) · consistency 61.1

raleigh, nc 80%
pittsburgh, pa 20%

openai/gpt-6-astra (10 runs) · consistency 58.3

raleigh, nc 80%
madison, wi 10%
minneapolis, mn 10%

openai/gpt-5.6-luna (20 runs) · consistency 58.3

raleigh, nc 80%
madison, wi 10%

openai/gpt-5.6-sol (20 runs) · consistency 77.8

raleigh, nc 90%
pittsburgh, pa 10%

openai/gpt-5.6-terra (20 runs) · consistency 58.3

raleigh, nc 75%
madison, wi 20%

openai/gpt-5.5 (10 runs) · consistency 100

raleigh, nc 100%

openai/gpt-5.4-mini (15 runs) · consistency 44.4

austin, tx 73%
raleigh, nc 13%
madison, wi 13%

openai/gpt-5.4-nano (30 runs) · consistency 33.3

austin, tx 33%
hedge 13%
seattle, wa 30%
boston, ma 10%

openai/gpt-5.4 (30 runs) · consistency 13.9

austin, tx 13%
raleigh, nc 13%
madison, wi 37%
chicago, il 13%
minneapolis, mn 10%

openai/gpt-5.3-chat (25 runs) · consistency 19.4

austin, tx 44%
raleigh, nc 28%
hedge 20%

openai/gpt-oss-120b (15 runs) · consistency 61.1

austin, tx 73%
other 20%

openai/o3 (10 runs) · consistency 77.8

austin, tx 90%
boulder, co 10%

openai/o4-mini (10 runs) · consistency 77.8

austin, tx 90%
raleigh, nc 10%

openai/gpt-4.1 (10 runs) · consistency 100

austin, tx 100%

openai/gpt-4.1-mini (20 runs) · consistency 44.4

austin, tx 70%
hedge 25%

openai/gpt-4.1-nano (10 runs) · consistency 77.8

austin, tx 90%
hedge 10%

openai/o3-mini (10 runs) · consistency 100

austin, tx 100%

openai/gpt-4o-mini (10 runs) · consistency 100

austin, tx 100%

openai/gpt-3.5-turbo (10 runs) · consistency 30.6

austin, tx 30%
hedge 10%
refusal 50%
seattle, wa 10%

openai/gpt-4 (10 runs) · consistency 30.6

austin, tx 60%
boulder, co 10%
hedge 20%
other 10%

Qwen

qwen/qwen3.8-max-0902 (10 runs) · consistency 44.4

austin, tx 70%
raleigh, nc 20%
huntsville, al 10%

qwen/qwen3.8-flash (5 runs)

austin, tx 80%
boulder, co 20%

qwen/qwen3.8-27b (6 runs)

austin, tx 83%
denver, co 17%

qwen/qwen3.7-flash (9 runs) · consistency 19.4

austin, tx 45%
raleigh, nc 22%
madison, wi 11%
refusal 11%
san diego, ca 11%

qwen/qwen3.7-plus (25 runs) · consistency 27.8

austin, tx 40%
raleigh, nc 24%
other 16%
huntsville, al 16%

qwen/qwen3.7-max (15 runs) · consistency 44.4

raleigh, nc 73%
green bay, wi 20%

qwen/qwen3.6-27b (20 runs) · consistency 36.1

austin, tx 65%
raleigh, nc 20%
refusal 10%

qwen/qwen3.6-flash (20 runs) · consistency 36.1

austin, tx 55%
raleigh, nc 35%

qwen/qwen3.6-max-preview (20 runs) · consistency 22.2

austin, tx 50%
raleigh, nc 20%
boulder, co 10%
refusal 20%

qwen/qwen3.6-plus (10 runs) · consistency 100

austin, tx 100%

qwen/qwen3.5-122b-a10b (25 runs) · consistency 27.8

austin, tx 40%
raleigh, nc 44%
boulder, co 16%

qwen/qwen3.5-flash-02-23 (30 runs) · consistency 30.6

austin, tx 27%
raleigh, nc 40%
hedge 27%

qwen/qwen3-235b-a22b-2507 (15 runs) · consistency 61.1

austin, tx 73%
refusal 27%

Sakana

sakana/fugu-ultra (15 runs) · consistency 77.8

raleigh, nc 93%

StepFun

stepfun/step-3.7-flash (20 runs) · consistency 25

austin, tx 65%
hedge 20%

Tencent

tencent/hy4-preview (4 runs)

raleigh, nc 100%

tencent/hy-mt2-30b-a3b (10 runs) · consistency 100

austin, tx 100%

tencent/hy3 (10 runs) · consistency 44.4

austin, tx 70%
denver, co 10%
boulder, co 20%

tencent/hy3:free (20 runs) · consistency 58.3

austin, tx 75%
raleigh, nc 10%

ThinkingMachines

thinkingmachines/inkling-small (10 runs) · consistency 25

austin, tx 50%
raleigh, nc 10%
madison, wi 10%
hedge 30%

thinkingmachines/inkling (20 runs) · consistency 36.1

austin, tx 70%
hedge 25%

Upstage

upstage/solar-mini4 (10 runs) · consistency 41.7

austin, tx 20%
hedge 60%
seattle, wa 10%
oregon city, or 10%

upstage/solar-pro4 (10 runs) · consistency 25

austin, tx 40%
hedge 40%
refusal 10%
seattle, wa 10%

upstage/solar-pro-3 (10 runs) · consistency 30.6

austin, tx 60%
raleigh, nc 10%
hedge 20%
dallas, tx 10%

xAI

x-ai/grok-4.7 (10 runs) · consistency 50

austin, tx 70%
raleigh, nc 30%

x-ai/grok-4.6 (10 runs) · consistency 100

austin, tx 100%

x-ai/grok-4.5 (20 runs) · consistency 58.3

austin, tx 90%

x-ai/grok-4.3 (20 runs) · consistency 30.6

austin, tx 60%
denver, co 10%
refusal 15%
other 15%

x-ai/grok-4.20 (20 runs) · consistency 77.8

austin, tx 90%

x-ai/grok-4-fast (10 runs) · consistency 100

austin, tx 100%

x-ai/grok-4.1-fast (10 runs) · consistency 100

austin, tx 100%

Xiaomi

xiaomi/mimo-v2.5 (20 runs) · consistency 77.8

austin, tx 80%
raleigh, nc 10%

xiaomi/mimo-v2.5-pro (20 runs) · consistency 77.8

austin, tx 90%

xiaomi/mimo-v2-omni (20 runs) · consistency 50

austin, tx 40%
raleigh, nc 60%

xiaomi/mimo-v2-pro (15 runs) · consistency 61.1

austin, tx 80%
raleigh, nc 20%

Z.ai

z-ai/glm-5.3-flash (10 runs) · consistency 44.4

austin, tx 50%
raleigh, nc 50%

z-ai/glm-5.3 (10 runs) · consistency 50

austin, tx 40%
raleigh, nc 60%

z-ai/glm-5.2 (15 runs) · consistency 50

austin, tx 67%
raleigh, nc 33%

z-ai/glm-5.1 (15 runs) · consistency 36.1

austin, tx 67%
raleigh, nc 27%

z-ai/glm-5-turbo (10 runs) · consistency 25

austin, tx 40%
raleigh, nc 40%
hedge 10%
arlington, va 10%

z-ai/glm-5 (20 runs) · consistency 44.4

austin, tx 30%
hedge 55%
other 10%

z-ai/glm-4.7-flash (20 runs) · consistency 77.8

austin, tx 10%
hedge 75%

z-ai/glm-4.7 (20 runs) · consistency 19.4

austin, tx 35%
hedge 20%
green bay, wi 35%