AI coding assistants for small business buyers
Asked as “AI coding assistant”, and as “AI pair programming tool”, on behalf of a small B2B company. 43 first choices recorded across the direct, paraphrase, budget and scale prompts, twelve models each.
Standing · first-choice share
65%
Clear leader
65GitHub Copilot19Windsurf09Cursor07others
65% of first choices, clear leader.
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.
The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all twelve models, for a small B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
Show the two products at 0%, ordered by negative rate
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.
Claude Code is made by Anthropic, whose model Claude Haiku 4.5 is in the set. On the four share prompts that model made it the first choice zero times of 4; the other eleven models zero times of 44. Lab treatment is defined on the method page; the row is marked, not excluded.
Head to headHead-to-head pages are published for mid-market buyers, so these open on mid-market figures, not small business.
All fifteen head-to-head pages: the top six products, each against each
One product takes 65% of first choices here, so the chart would put five 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. One product carries a negative rate above 25% in this category.
What they warned about
Seven of twelve models held their first choice under the paraphrase. Claude Haiku 4.5, Gemini 3.5 Flash, DeepSeek V4 Flash, Llama 4 Maverick and GLM 4.7 FlashX 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.
GitHub Copilot14%
10 of 71 labels negative · 8 of 12 models · 4 hard negative
“Free, Pro, Pro+ plans use your code, prompts, and context for training models by default... risk IP exposure to competitors.” Grok 4.20, negative prompt
Cursor17%
10 of 60 labels negative · 8 of 12 models · 2 hard negative
“Privacy Mode (no training on data) is off by default for non-enterprise. Recent IDEsaster flaws (CVE-2025-49150) allow prompt injection for data exfil/RCE.” Grok 4.20, negative prompt
Windsurf12%
6 of 49 labels negative · 6 of 12 models · 1 hard negative
“Enterprise version safer, but free appeals to cost-conscious SMBs—avoid without verifying opt-outs.” Grok 4.20, negative prompt
Amazon Q Developer25%
6 of 24 labels negative · 5 of 12 models · 2 hard negative
“Any Unvetted Free/Open AI Tools (e.g., lesser-known like Amazon Q Developer Free, Replit AI): High malware/fake tool risks” Grok 4.20, negative prompt
What they cite
Twelve of the twelve models returned a source list in this edition: four return the pages they cited and eight the search results they consulted.
Sites the answers cite
64 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. 905 links across 343 sites, every framing counted. Ranked by the number of answers carrying the site or page.
24 answers · 24 citations · 10 models
21 answers · 25 citations · 9 models
19 answers · 21 citations · 8 models
vendor site · G219 answers · 19 citations · 9 models
vendor site · Zapier17 answers · 17 citations · 10 models
15 answers · 15 citations · 8 models
vendor site · G214 answers · 14 citations · 9 models
13 answers · 13 citations · 7 models
vendor site · Augment Code12 answers · 15 citations · 8 models
vendor site · Daily12 answers · 15 citations · 6 models
12 answers · 12 citations · 7 models
11 answers · 13 citations · 9 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.
24 answers · 10 models
19 answers · 9 models
17 answers · 10 models
16 answers · 7 models
15 answers · 8 models
12 answers · 7 models
12 answers · 9 models
10 answers · 5 models
8 answers · 5 models
8 answers · 8 models
Search against answers
Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “ai coding assistant”, “ai coding assistants”, “best ai coding assistant”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
| Company and products | Share | Own site cited | Name searches, Google | AI search demand, est. | Organic visits | Paid search a month, est. | Google ads, last 30 days | Best Google position |
|---|---|---|---|---|---|---|---|---|
| GitHub github.com GitHub Copilot 65% | 65% | 207 | 1,500,000 | 31,479 | 42,019,137 company | $81 | 14 | |
| Windsurf windsurf.com | 19% | 66 | 60,500 | 4,501 | 21,603 | 0 | ||
| Cursor cursor.com | 9% | 164 | 301,000 | 22,500 | 350,718 | $28 | 119 | 6 |
| Amazon Q amazon.com Amazon Q Developer 5% | 5% | 109 | 9,900 | 738 | 1,681,997,879 company | $106,125 | 115 | |
| Continue continue.dev | 2% | 19 | 673,000 | 70,608 | 2,964 | |||
| Tabnine tabnine.com | 0% | 89 | 2,900 | 62 | 15,577 | 0 | ||
| Claude Code claude.com | 0% | 92 | 550,000 | 9,804 | 3,090,945 | $1,889 | 120 | |
| Cline cline.bot | 0% | 23 | 22,200 | 8,724 | 12,658 | 3 | ||
| Aider aider.chat | 0% | 20 | 9,900 | 585 | 5,507 | |||
| Sourcegraph Cody sourcegraph.com | 0% | 30 | 880 | 14 | 10,627 | 14 |
Google's top ten for the three searches hold 24 sites; 2 of them are among the sites the models cited here (augmentcode.com, axify.io). The first result for “ai coding assistant” is augmentcode.com.
How 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.ShowHide
How 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.
ShowHideModelDirect“What is the best AI coding assistant for a small B2B company?”Paraphrase“Which AI pair programming tool would you recommend to a small business that sells to other businesses?”Comparative“What are the top AI coding assistants for a small team and how do they differ?”Budget-constrained“What is the best AI coding assistant for a small company with a tight budget?”Scale-constrained“We are a 30 person company evaluating an AI coding assistant. What should we look at?”Negative“Which AI coding assistants should a small business avoid or be cautious about?”
Claude Haiku 4.5no first choiceGitHub CopilotChangedCursorGitHub Copilotno first choicenothing named
One alternative
CursorFour alternatives
Claude Code, GitHub Copilot, Tabnine, WindsurfTwo alternatives
Amazon Q Developer, CursorGPT-5.4 miniGitHub CopilotGitHub CopilotHeldGitHub CopilotGitHub Copilotno first choicenothing named
Two alternatives
Cursor, OpenAI CodexOne alternative
Microsoft 365 Copilot for BusinessThree alternatives
Cursor, Gemini Code Assist, JetBrains AI AssistantOne alternative
CursorGemini 3.5 FlashCursorGitHub CopilotChangedagainst: ChatGPTCursorContinueCursor, GitHub Copilotagainst: ChatGPT, Claude, Claude Code, Cursor, Google Gemini, Windsurf
Three alternatives
Amazon Q Developer, GitHub Copilot, TabnineTwo alternatives
Cursor, TabnineFour alternatives
Claude Code, Continue, Devin, GitHub CopilotFour alternatives
Cursor, DeepSeek, GitHub Copilot, WindsurfTwo alternatives
Continue, TabninePerplexity SonarGitHub CopilotGitHub CopilotHeldCursor, GitHub CopilotGitHub Copilotno first choicenothing named
Three alternatives
Claude Code, Cline, CursorTwo alternatives
Cursor, WindsurfTwo alternatives
Claude Code, TabnineThree alternatives
Aider, Cursor, WindsurfGrok 4.20GitHub CopilotGitHub CopilotHeldagainst: CursorCursor, GitHub CopilotWindsurfagainst: CursorCursor, GitHub Copilotagainst: Amazon Q Developer, Cursor, GitHub Copilot, Replit AI, Windsurf, Zed
Two alternatives
Cursor, WindsurfOne alternative
WindsurfThree alternatives
Amazon Q Developer, Tabnine, WindsurfOne alternative
GitHub CopilotTwo alternatives
Tabnine, WindsurfMistral SmallGitHub CopilotGitHub CopilotHeldGitHub CopilotGitHub Copilotno first choiceagainst: GitHub Copilot
One alternative
CursorTwo alternatives
Cursor, WindsurfFour alternatives
Claude Code, Cursor, Replit AI, WindsurfTwo alternatives
Cline, WindsurfDeepSeek V4 FlashWindsurfGitHub CopilotChangedGitHub Copilotagainst: WindsurfGitHub Copilotagainst: Amazon Q Developer, Gemini Code Assistno first choiceagainst: Bolt, Claude Code, Cursor, GitHub Copilot, Lovable, Replit AI, v0
Three alternatives
Cursor, GitHub Copilot, TabnineOne alternative
TabnineThree alternatives
Claude Code, Cursor, TabnineTwo alternatives
Cursor, WindsurfLlama 4 MaverickGitHub CopilotCursor, WindsurfChangedCursorGitHub Copilotno first choiceagainst: Cursor, GitHub Copilot, Windsurf
One alternative
CursorOne alternative
GitHub CopilotFour alternatives
Cline, GitHub Copilot, Kilo Code, Qwen3-CoderTwo alternatives
Cline, WindsurfFour alternatives
Amazon Q Developer, Claude Code, Cursor, GitHub CopilotQwen 3.7 FlashGitHub Copilotagainst: Amazon Q DeveloperGitHub CopilotHeldCursor, GitHub CopilotAmazon Q Developer, Windsurfno first choiceagainst: Amazon Kiro, Amazon Q Developer, Bolt.new, Claude Code, Cursor, Devin, GitHub Copilot, Google Antigravity IDE, Loveable, Replit Agent, Tabnine, Windsurf, v0.dev
Two alternatives
Cursor, TabnineOne alternative
TabnineTwo alternatives
Sourcegraph Cody, TabnineTwo alternatives
Cursor, GitHub CopilotKimi K2GitHub CopilotGitHub CopilotHeldCursor, GitHub CopilotWindsurfGitHub Copilotagainst: Amazon Q Developer, ChatGPT, Cursor, GitHub Copilot
Two alternatives
Cursor, TabnineOne alternative
CursorThree alternatives
Claude Code, Tabnine, WindsurfTwo alternatives
GitHub Copilot, TabnineTwo alternatives
Cursor, WindsurfGLM 4.7 FlashXGitHub CopilotWindsurfChangedGitHub CopilotAmazon Q Developer, Windsurfagainst: GitHub Copilot, Tabnineno first choiceagainst: Amazon Q Developer, Claude Code, Cursor, GitHub Copilot, Trae
Five alternatives
Amazon Q Developer, Claude Code, Cursor, Tabnine, WindsurfTwo alternatives
GitHub Copilot, TabnineFive alternatives
Cursor, Replit AI, Sourcegraph Cody, Tabnine, WindsurfOne alternative
Continue.dev with OllamaMiniMax M2.5GitHub CopilotGitHub CopilotHeldGitHub CopilotWindsurfno first choiceagainst: ChatGPT, Claude, Cursor, GitHub Copilot, Windsurf
One alternative
CursorThree alternatives
Amazon Q Developer, JetBrains AI Assistant, TabnineThree alternatives
Claude Code, Cursor, WindsurfTwo alternatives
GitHub Copilot, TabnineBold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency.
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 08:20 | no | 0 | 4 s |
| Direct recommendation | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 08:19 | yes | 3 | 6 s |
| Direct recommendation | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 09:06 | yes | 10 | 22 s |
| Direct recommendation | Perplexity Sonar | sonar | 2026-09-17 08:21 | yes | 20 | 5 s |
| Direct recommendation | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 08:17 | yes | 17 | 9 s |
| Direct recommendation | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 08:49 | yes | 5 | 4 s |
| Direct recommendation | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 09:11 | yes | 18 | 16 s |
| Direct recommendation | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 08:48 | yes | 5 | 3 s |
| Direct recommendation | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 08:44 | yes | 10 | 60 s |
| Direct recommendation | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:32 | yes | 19 | 31 s |
| Direct recommendation | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:24 | yes | 10 | 46 s |
| Direct recommendation | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 08:38 | yes | 8 | 18 s |
| Paraphrase | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-09-17 08:47 | yes | 18 | 9 s |
| Paraphrase | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 09:04 | yes | 2 | 4 s |
| Paraphrase | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 09:02 | yes | 19 | 22 s |
| Paraphrase | Perplexity Sonar | sonar | 2026-09-17 09:00 | yes | 20 | 5 s |
| Paraphrase | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 09:01 | yes | 41 | 16 s |
| Paraphrase | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 08:56 | yes | 5 | 4 s |
| Paraphrase | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 08:36 | yes | 20 | 19 s |
| Paraphrase | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 08:24 | yes | 5 | 3 s |
| Paraphrase | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 08:23 | yes | 5 | 32 s |
| Paraphrase | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:36 | yes | 15 | 24 s |
| Paraphrase | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:43 | yes | 25 | 36 s |
| Paraphrase | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 08:51 | yes | 5 | 67 s |
| Comparative | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-09-17 09:09 | yes | 9 | 10 s |
| Comparative | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 08:34 | yes | 4 | 7 s |
| Comparative | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 09:08 | yes | 21 | 28 s |
| Comparative | Perplexity Sonar | sonar | 2026-09-17 08:42 | yes | 20 | 8 s |
| Comparative | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 09:03 | yes | 46 | 17 s |
| Comparative | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 08:37 | yes | 5 | 7 s |
| Comparative | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 08:55 | yes | 20 | 20 s |
| Comparative | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 09:06 | yes | 5 | 2 s |
| Comparative | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 08:50 | no | 0 | 44 s |
| Comparative | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:23 | yes | 10 | 35 s |
| Comparative | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:20 | yes | 10 | 66 s |
| Comparative | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 08:48 | yes | 5 | 21 s |
| Budget constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-09-17 08:54 | no | 0 | 5 s |
| Budget constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 08:16 | yes | 3 | 6 s |
| Budget constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 08:59 | yes | 16 | 23 s |
| Budget constrained | Perplexity Sonar | sonar | 2026-09-17 08:19 | yes | 20 | 6 s |
| Budget constrained | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 08:46 | yes | 18 | 9 s |
| Budget constrained | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 08:30 | yes | 5 | 6 s |
| Budget constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 08:22 | yes | 9 | 16 s |
| Budget constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 08:19 | yes | 5 | 2 s |
| Budget constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 08:45 | no | 0 | 40 s |
| Budget constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:20 | yes | 10 | 22 s |
| Budget constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:47 | yes | 30 | 144 s |
| Budget constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 08:18 | yes | 10 | 16 s |
| Scale constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-09-17 08:46 | no | 0 | 5 s |
| Scale constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 08:51 | no | 0 | 7 s |
| Scale constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 08:21 | yes | 10 | 23 s |
| Scale constrained | Perplexity Sonar | sonar | 2026-09-17 08:26 | yes | 20 | 7 s |
| Scale constrained | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 09:04 | yes | 20 | 11 s |
| Scale constrained | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 09:10 | no | 0 | 6 s |
| Scale constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 08:20 | yes | 15 | 18 s |
| Scale constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 08:44 | yes | 5 | 4 s |
| Scale constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 09:09 | no | 0 | 40 s |
| Scale constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:27 | yes | 15 | 39 s |
| Scale constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:45 | yes | 15 | 115 s |
| Scale constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 08:27 | yes | 5 | 28 s |
| Negative framing | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-09-17 09:12 | yes | 18 | 10 s |
| Negative framing | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-09-17 08:39 | yes | 3 | 11 s |
| Negative framing | Gemini 3.5 Flash | gemini-3.5-flash | 2026-09-17 08:42 | yes | 14 | 26 s |
| Negative framing | Perplexity Sonar | sonar | 2026-09-17 08:24 | yes | 20 | 6 s |
| Negative framing | Grok 4.20 | spacexai/grok-4.1-fast-non-reasoning via vertex (asked for grok-4.20-non-reasoning; the API answered as this version) | 2026-09-17 09:06 | yes | 43 | 16 s |
| Negative framing | Mistral Small | mistral/mistral-small via mistral | 2026-09-17 09:12 | yes | 5 | 6 s |
| Negative framing | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via fireworks | 2026-09-17 09:10 | yes | 26 | 28 s |
| Negative framing | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-09-17 08:58 | yes | 5 | 2 s |
| Negative framing | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-09-17 08:30 | yes | 19 | 92 s |
| Negative framing | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-09-17 08:58 | yes | 19 | 70 s |
| Negative framing | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-09-17 08:41 | yes | 29 | 53 s |
| Negative framing | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-09-17 09:05 | yes | 13 | 44 s |
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
ShowHide
Category-scoped readings
Amazon Q read as Amazon Q Developer
Copilot Free read as GitHub Copilot
JetBrains AI read as JetBrains AI Assistant
Replit read as Replit AI
Unresolved, counted raw
Amazon Kiro
Base44
Bolt
CodeLlama
Continue.dev with Ollama
Loveable
Microsoft 365 Copilot for Business
Open-source LLMs (self-hosted)
v0
v0.dev
Discontinued, still offered
No shut-down product was recommended here.