Index › Developer platform › AI coding › Enterprise › September 2026 Edition

AI coding assistants for enterprise buyers

Asked as “AI coding assistant”, and as “AI pair programming tool”, on behalf of an enterprise B2B company. 37 first choices recorded across the direct, paraphrase, budget and scale prompts, twelve models each.

Standing

GitHub Copilot

Consensus

95% of first choices, consensus.

ShowinginAll 67 categories →

By buyer segment

The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.

Small business

GitHub Copilot65%Clear leaderStanding →Mid-market

GitHub Copilot67%Clear leaderStanding →Enterprise

GitHub Copilot95%ConsensusThis page

01The standing

Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all twelve models, for an enterprise B2B company. Ordered by share.

ProductFirst-choice shareNegative rateLabelsQuadrant

GitHub Copilot16%68endorsed leader

Amazon Q Developer12%40accepted challenger

Show the seven products at 0%, ordered by negative rate

Cursor45%49criticized challenger

Claude Code40%25criticized challenger

Windsurf21%14accepted challenger

Gemini Code Assist17%12accepted challenger

Tabnine7%44accepted challenger

Augment Code0%16accepted challenger

Sourcegraph Cody0%14accepted challenger

Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its vendor page.

Head to head

GitHub Copilot vs Windsurf GitHub Copilot vs Cursor Windsurf vs Cursor

One product takes 95% of first choices here, so the chart would put eight markers in one corner and one at the far edge. The two measurements it plots are columns in the standing above: share, and the negative label rate. Two products carry a negative rate above 25% in this category. Jump to the standing

02What they warned about

Eleven of twelve models held their first choice under the paraphrase. Llama 4 Maverick changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.

Cursor

45%

22 of 49 labels negative · 10 of 12 models · 4 hard negative

“Outright Avoid: Consumer/Free Tiers of Major Tools ... the free tier has no contractual zero-retention guarantee” DeepSeek V4 Flash, negative prompt

GitHub Copilot

16%

11 of 68 labels negative · 9 of 12 models · 4 hard negative

“interaction data ... will be used to **train their AI models by default** ... **Enterprise action:** Ban personal account usage for work code immediately.” DeepSeek V4 Flash, negative prompt

Claude Code

40%

10 of 25 labels negative · 6 of 12 models · 4 hard negative

“High Caution / Avoid for Sensitive Environments ... Exercise extreme caution; assess geopolitical exposure” Kimi K2, negative prompt

Amazon Q Developer

12%

5 of 40 labels negative · 5 of 12 models · 1 hard negative

“avoid cloud-only solutions such as **GitHub Copilot, Amazon Q Developer, and Cursor** in air-gapped scenarios” Perplexity Sonar, negative prompt

03What they cite

Citations exist only for the models that return a source list: twelve of the twelve in this edition, and all six flagship models on the expanded tier.

Sites the answers cite

67 of 72 answers in this category came back with a source list, from 12 of 12 models: citations where the model returns them, or the search results it consulted. 5 of those lists are Google grounding redirects that name no site and are left out of the counts. 871 links across 307 sites, every framing counted. Ranked by the number of answers carrying the site or page.

augmentcode.com

vendor site · Augment Code39 answers · 84 citations · 10 models

axify.io

23 answers · 23 citations · 10 models

agentic.ai

15 answers · 15 citations · 9 models

blog.exceeds.ai

15 answers · 15 citations · 8 models

intuitionlabs.ai

14 answers · 15 citations · 6 models

github.com

vendor site · GitHub12 answers · 17 citations · 6 models

learn.g2.com

12 answers · 12 citations · 6 models

tabnine.com

vendor site · Tabnine11 answers · 22 citations · 6 models

baeseokjae.github.io

vendor site · GitHub11 answers · 13 citations · 8 models

omidsaffari.com

11 answers · 11 citations · 7 models

docs.github.com

vendor site · GitHub10 answers · 23 citations · 5 models

aws.amazon.com

vendor site · Amazon10 answers · 20 citations · 5 models

Pages the answers cite

The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.

axify.io/blog/the-best-ai-coding-assistants-a-full-comparison-of-20-tools

23 answers · 10 models

augmentcode.com/tools/beyond-github-copilot-5-enterprise-ready-ai-coding-as…

18 answers · 8 models

agentic.ai/best/enterprise-coding-agents

15 answers · 9 models

augmentcode.com/tools/8-top-ai-coding-assistants-and-their-best-use-cases

12 answers · 7 models

learn.g2.com/best-ai-coding-assistants

12 answers · 6 models

omidsaffari.com/blog/best-ai-coding-agents-for-enterprise

11 answers · 7 models

witness.ai/blog/ai-coding-assistants-security

10 answers · 8 models

getdx.com/blog/ai-coding-assistant-pricing

9 answers · 8 models

wwt.com/wwt-research/how-to-securely-implement-ai-coding-assistants-across-…

9 answers · 8 models

docs.github.com/en/copilot/get-started/plans

8 answers · 4 models

04How they answered

Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.

ModelDirect“What is the best AI coding assistant for an enterprise B2B company?”Paraphrase“Which AI pair programming tool would you recommend to a large B2B company with thousands of employees?”Comparative“What are the top enterprise-grade AI coding assistants and how do they differ?”Budget-constrained“What is the best AI coding assistant for a large company that needs predictable total cost across thousands of users?”Scale-constrained“We are a 5,000 person company with SSO, SOC 2 and procurement review requirements evaluating an AI coding assistant. What should we look at?”Negative“Which AI coding assistants should a large enterprise avoid or be cautious about?”

Claude Haiku 4.5GitHub Copilot

Four alternatives

Claude, Cursor, Tabnine, Windsurf

GitHub CopilotHeld

Three alternatives

Amazon Q Developer, JetBrains AI Assistant, Tabnine

no first choice

Five alternatives

Amazon Q Developer, Augment Code, Cursor, GitHub Copilot, Tabnine

GitHub Copilot

Two alternatives

Cursor, Tabnine

no first choiceagainst: ChatGPT, Claude, Cursor, Google Gemini

GPT-5.4 miniGitHub Copilot

Three alternatives

Amazon Q Developer, Gemini Code Assist, OpenAI Codex / ChatGPT Enterprise + Codex

GitHub CopilotHeld

Two alternatives

Amazon Q Developer, Gemini Code Assist

GitHub Copilot

Three alternatives

Amazon Q Developer, Cursor, Gemini Code Assist

GitHub Copilot

Two alternatives

Amazon Q Developer, Tabnine

no first choiceagainst: Cursoragainst: Cursor

Gemini 3.5 FlashGitHub Copilot

Four alternatives

Cursor, Sourcegraph Cody, Tabnine, Windsurf

GitHub CopilotHeld

Four alternatives

Amazon Q Developer, Augment Code, Sourcegraph Cody, Tabnine

against: Claude Code, Cursorno first choice

Six alternatives

Augment Code, Claude Code, Cursor, GitHub Copilot, Sourcegraph Cody, Tabnine

GitHub Copilot

Two alternatives

Cursor, Tabnine

no first choiceagainst: Aider, Blackbox AI, ChatGPT, Claude, Claude Code, ClawHub / OpenClaw, Cursor, GitHub Copilot, Supermaven

Perplexity SonarGitHub Copilot

Five alternatives

Amazon Q Developer, Augment Code, Gemini Code Assist, Sourcegraph Cody, Tabnine

GitHub CopilotHeld

Two alternatives

Gemini Code Assist, Tabnine

against: CursorGitHub Copilot

Seven alternatives

Amazon Q Developer, Augment Code, Claude Code, Cursor, Gemini Code Assist, OpenAI Codex, Tabnine

GitHub Copilotagainst: Claude Code, Cursorno first choiceagainst: Amazon Q Developer, ChatGPT, Claude Code, Cursor, GitHub Copilot, JetBrains Junie, OpenAI Codex, Roo Code

Grok 4.1 FastGitHub Copilot

Four alternatives

Amazon Q Developer, Cursor, Tabnine, Windsurf

GitHub CopilotHeld

Three alternatives

Amazon Q Developer, Cursor, Tabnine

no first choiceGitHub Copilot

Two alternatives

Amazon Q Developer, Tabnine

against: Claude Code, Cursor, Gemini Code Assist, Windsurf, raw OpenAIGitHub Copilot

One alternative

Amazon Q Developer

against: Augment, Cursor, Replit Agentagainst: Amazon Q Developer, Cursor, Gemini Code Assist, GitHub Copilot, Tabnine

Mistral SmallGitHub Copilot

Two alternatives

Claude Code, Cursor

GitHub CopilotHeld

Three alternatives

JetBrains AI Assistant, Tabnine, Zed

GitHub Copilot

Four alternatives

Amazon Q Developer, Augment Code, Claude Code, Cursor

GitHub Copilot

One alternative

Tabnine

no first choicenothing named

DeepSeek V4 FlashGitHub Copilot

Two alternatives

Claude Code, OpenAI Codex

against: Cursor, WindsurfGitHub CopilotHeld

Two alternatives

Amazon Q Developer, Tabnine

against: CursorGitHub Copilot

Five alternatives

Amazon Q Developer, Augment Code, Cursor, Sourcegraph Cody, Tabnine

GitHub Copilot

Three alternatives

Amazon Q Developer, Gemini Code Assist, Tabnine

against: Claude Code, Cursor, OpenAI Codexno first choiceagainst: Cursoragainst: Claude Code, Cursor, GitHub Copilot, OpenAI Codex, Windsurf

Llama 4 MaverickGitHub Copilot

Three alternatives

Claude Code, Cursor, OpenAI Codex

no first choiceChangedno first choiceGitHub Copilot

Two alternatives

Claude Code, Cursor

no first choiceagainst: Amazon Q Developer, ChatGPT, GitHub Copilot

Qwen 3.7 FlashGitHub Copilot

Three alternatives

Amazon Q Developer, Cursor, Tabnine

GitHub CopilotHeld

Two alternatives

Amazon Q Developer, Tabnine

GitHub Copilot

Six alternatives

Amazon Q Developer, Augment Code, Cursor, JetBrains AI Assistant, Sourcegraph Cody, Tabnine

GitHub Copilot

Two alternatives

JetBrains AI Assistant, Tabnine

GitHub Copilot

Five alternatives

Cursor, Exceeds AI, Kilo Code, Tabnine, Windsurf

against: Amazon Q Developer, Claude Code, Cursor, GitHub Copilot, OpenClaw

Kimi K2GitHub Copilot

Two alternatives

Amazon Q Developer, Tabnine

GitHub CopilotHeld

Two alternatives

Amazon Q Developer, Tabnine

GitHub Copilot

Three alternatives

Amazon Q Developer, Sourcegraph Cody, Tabnine

against: CursorAmazon Q Developer

Two alternatives

GitHub Copilot, Tabnine

against: Claude Code, Cursorno first choiceagainst: Amazon Q Developer, Claude Code, GitHub Copilot

GLM 4.7 FlashXGitHub Copilot

Seven alternatives

Amazon Q Developer, Augment Code, Claude Code, Cursor, JetBrains AI Assistant, Sourcegraph Cody, Tabnine

GitHub CopilotHeld

Three alternatives

Amazon Q Developer, JetBrains AI, Sourcegraph Cody

GitHub Copilot

Five alternatives

Amazon Q Developer, Augment Code, JetBrains AI Assistant, Sourcegraph Cody, Tabnine

Amazon Q Developer

One alternative

JetBrains AI + Anthropic/Claude

against: Cursor, GitHub Copilot, Tabnineno first choicenothing named

MiniMax M2.5GitHub Copilot

Three alternatives

Augment Code, Claude Code, Cursor

GitHub CopilotHeld

Two alternatives

Amazon Q Developer, Tabnine

no first choiceGitHub Copilotagainst: Cursor, Tabnineno first choiceagainst: ChatGPT, Claude, Cursor, GitHub Copilot, Google's Gemini-based tools

Bold is the first choiceAlternatives are counted; the count opens them.What the answer argued against

05The record

One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record

Seventy-two rows: every prompt, every model, every answer.

Prompt	Model	Version string	Time (UTC)	Searched	Sources	Latency
Direct recommendation	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 09:08	yes	18	9 s
Direct recommendation	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 08:23	yes	4	7 s
Direct recommendation	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 09:09	yes	13	26 s
Direct recommendation	Perplexity Sonar	`sonar`	2026-09-17 08:51	yes	20	5 s
Direct recommendation	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 08:27	yes	45	18 s
Direct recommendation	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 09:10	yes	5	4 s
Direct recommendation	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 08:58	yes	10	19 s
Direct recommendation	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 09:09	yes	5	2 s
Direct recommendation	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 08:55	yes	5	31 s
Direct recommendation	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 08:32	yes	10	51 s
Direct recommendation	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:35	yes	12	88 s
Direct recommendation	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 08:47	yes	5	13 s
Paraphrase	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 09:07	no	0	5 s
Paraphrase	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 09:03	yes	4	6 s
Paraphrase	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 09:12	yes	15	26 s
Paraphrase	Perplexity Sonar	`sonar`	2026-09-17 08:29	yes	19	6 s
Paraphrase	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 08:28	yes	21	10 s
Paraphrase	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 08:18	yes	5	6 s
Paraphrase	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 08:40	yes	10	12 s
Paraphrase	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 08:50	yes	5	3 s
Paraphrase	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 08:22	yes	5	32 s
Paraphrase	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 08:22	yes	25	28 s
Paraphrase	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:40	yes	25	79 s
Paraphrase	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 08:22	yes	10	50 s
Comparative	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 08:49	yes	9	11 s
Comparative	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 08:26	yes	8	7 s
Comparative	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 09:15	yes	20	29 s
Comparative	Perplexity Sonar	`sonar`	2026-09-17 09:10	yes	20	11 s
Comparative	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 08:49	yes	46	17 s
Comparative	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 08:44	yes	25	15 s
Comparative	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 08:56	yes	24	23 s
Comparative	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 08:18	yes	5	2 s
Comparative	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 08:39	yes	5	41 s
Comparative	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 08:44	yes	15	31 s
Comparative	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:34	yes	23	182 s
Comparative	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 08:31	yes	5	13 s
Budget constrained	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 08:36	yes	18	10 s
Budget constrained	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 08:35	yes	3	6 s
Budget constrained	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 08:59	yes	30	31 s
Budget constrained	Perplexity Sonar	`sonar`	2026-09-17 08:26	yes	20	6 s
Budget constrained	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 08:56	yes	26	15 s
Budget constrained	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 09:06	yes	5	3 s
Budget constrained	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 09:09	yes	10	18 s
Budget constrained	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 09:06	yes	5	3 s
Budget constrained	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 08:58	no	0	40 s
Budget constrained	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 08:42	yes	9	48 s
Budget constrained	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:18	yes	13	120 s
Budget constrained	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 08:50	yes	20	71 s
Scale constrained	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 08:42	yes	18	17 s
Scale constrained	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 08:22	yes	5	14 s
Scale constrained	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 08:29	no	0	17 s
Scale constrained	Perplexity Sonar	`sonar`	2026-09-17 08:21	yes	20	8 s
Scale constrained	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 08:23	yes	19	12 s
Scale constrained	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 09:03	no	0	11 s
Scale constrained	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 08:16	yes	13	24 s
Scale constrained	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 08:19	yes	5	4 s
Scale constrained	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 08:46	yes	10	44 s
Scale constrained	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 08:58	yes	14	85 s
Scale constrained	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:45	yes	14	40 s
Scale constrained	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 09:03	no	0	77 s
Negative framing	Claude Haiku 4.5	`claude-haiku-4-5-20251001`	2026-09-17 08:57	yes	18	11 s
Negative framing	GPT-5.4 mini	`gpt-5.4-mini-2026-03-17`	2026-09-17 08:38	yes	4	6 s
Negative framing	Gemini 3.5 Flash	`gemini-3.5-flash`	2026-09-17 09:15	yes	28	25 s
Negative framing	Perplexity Sonar	`sonar`	2026-09-17 09:12	yes	20	9 s
Negative framing	Grok 4.1 Fast	`spacexai/grok-4.1-fast-non-reasoning via vertex`	2026-09-17 09:13	yes	24	13 s
Negative framing	Mistral Small	`mistral/mistral-small via mistral`	2026-09-17 08:56	yes	5	7 s
Negative framing	DeepSeek V4 Flash	`deepseek/deepseek-v4-flash via fireworks`	2026-09-17 08:34	yes	28	38 s
Negative framing	Llama 4 Maverick	`meta/llama-4-maverick via bedrock`	2026-09-17 08:38	yes	5	2 s
Negative framing	Qwen 3.7 Flash	`alibaba/qwen3.7-flash via alibaba`	2026-09-17 09:01	yes	9	70 s
Negative framing	Kimi K2	`moonshotai/kimi-k2 via novita`	2026-09-17 09:11	yes	30	44 s
Negative framing	GLM 4.7 FlashX	`zai/glm-4.7-flashx via zai`	2026-09-17 08:44	yes	14	41 s
Negative framing	MiniMax M2.5	`minimax/minimax-m2.5 via minimax`	2026-09-17 08:54	yes	14	40 s

Normalization in this category

Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.

Category-scoped readings

Amazon Q read as Amazon Q Developer

Replit read as Replit AI

Replit Enterprise read as Replit AI

Unresolved, counted raw

Augment

Blackbox AI

ClawHub / OpenClaw

Exceeds AI

Google's Gemini-based tools

IBM watsonx Code Assistant

JetBrains AI + Anthropic/Claude

OpenAI Codex / ChatGPT Enterprise + Codex

OpenClaw

SoftSpell

Supermaven

Unity Gateway

raw OpenAI

Discontinued, still offered

No shut-down product was recommended here.

AI coding assistants for enterprise buyers

By buyer segment

01The standing

02What they warned about

03What they cite

Sites the answers cite

Pages the answers cite

04How they answered

05The record

Normalization in this category

Also in developer platform