Cohere: Command A+ vs Google: Gemini 3.1 Flash Lite Preview
A side-by-side comparison of Cohere: Command A+ ($1.05 / 1M tokens) and Google: Gemini 3.1 Flash Lite Preview ($0.88 / 1M tokens) using current USD prices and capabilities from the api14 catalog. Both run through the same OpenAI-compatible API.
Specifications
| Cohere: Command A+ | Google: Gemini 3.1 Flash Lite Preview | |
|---|---|---|
| Provider | Cohere | |
| Input | $1.05 / 1M tokens | $0.88 / 1M tokens |
| Output | $5.25 / 1M tokens | $5.25 / 1M tokens |
| Cache read | $0.52 / 1M tokens | $0.09 / 1M tokens |
| Cache write | - | $0.29 / 1M tokens |
| Reasoning | - | $5.25 / 1M tokens |
| Audio input | - | $1.75 / 1M tokens |
| Image | - | $0.000001 / image |
| Web search | - | $49.00 / 1K searches |
| Context window | 192,000 | 1,048,576 |
| Max output | 64,000 | 65,536 |
| Input | text, image | text, image, video, file, audio |
| Output | text | text |
| Tool calling | ||
| Structured output | ||
| Reasoning control | ||
| Added | Sep 22, 2026 | Mar 3, 2026 |
Key differences
- Google: Gemini 3.1 Flash Lite Preview is 17% cheaper on input tokens.
- Google: Gemini 3.1 Flash Lite Preview reads 5.5× more context (1,048,576 tokens).
- A request with 1,000 input and 500 output tokens costs $0.00367 with Cohere: Command A+ and $0.00350 with Google: Gemini 3.1 Flash Lite Preview.
In depth
Cohere: Command A+ is a text model from Cohere, released on 2026-09-22. Google: Gemini 3.1 Flash Lite Preview is a text model from Google, released on 2026-03-03. Both are served through api14 as chat models with text output, so at the level of a single completion call they look similar: you send messages, you get text back. The differences show up in context size, what you can feed in, and the tuning knobs each provider exposes.
The clearest split is context and input modalities. Gemini 3.1 Flash Lite Preview takes a much larger context window, 1,048,576 tokens against 192,000 for Command A+. Gemini also accepts more input types: text, image, video, file and audio, while Command A+ accepts text and image only. Both return text only. On parameters, Command A+ supports frequency_penalty, presence_penalty, top_k, stop and seed alongside temperature, top_p, tools, structured_outputs, response_format, reasoning and include_reasoning. Gemini supports temperature, top_p, seed, tools, tool_choice, structured_outputs, response_format, reasoning, reasoning_effort and include_reasoning, but not frequency_penalty, presence_penalty, top_k or stop. If your prompt pipeline relies on repetition penalties or a stop sequence, that is a reason to look closely at which model you call. Gemini is the cheaper option on input pricing, while output pricing is about the same between the two.
For day to day work, pick based on what you are feeding the model. If you are summarizing or reasoning over very long documents, large codebases, or long transcripts, Gemini's larger context means fewer chunking workarounds. If you need to pass audio or video directly, Gemini is the only one of the two that accepts those modalities. If you are working with images plus text and want finer control over sampling behavior, Command A+ gives you penalties and top_k, which can matter for keeping generated text on track. Both support tools and structured outputs, so agent style and JSON shaped workloads are viable on either; neither has an exclusive advantage there according to the available data. For straightforward text generation where cost of input tokens dominates, Gemini is the cheaper call.
Calling either model through api14 uses the same OpenAI-compatible interface. You point your client at the api14 base URL, send a standard chat completions request with your messages array, and set the model field. To use Cohere's model, set model to "cohere/command-a-plus". To use Google's model, set model to "google/gemini-3.1-flash-lite-preview". Everything else in the request body stays the same, with one caveat: only pass parameters the chosen model supports. A request that includes top_k or stop will not carry over cleanly to the Gemini model, and reasoning_effort applies to Gemini but not to Command A+. Keep your parameter set to the common subset (temperature, top_p, max_tokens, tools, structured_outputs, response_format, seed, reasoning) if you want to switch model ids without touching anything else.
Switch with one string
Both models use the same endpoint and request shape. Change the model value from "cohere/command-a-plus" to "google/gemini-3.1-flash-lite-preview" and prompts, streaming and tools keep working.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api14.com/api/v1",
apiKey: process.env.API14_API_KEY,
});
const result = await client.chat.completions.create({
"model": "google/gemini-3.1-flash-lite-preview",
"messages": [
{
"role": "user",
"content": "Explain vector databases in two sentences."
}
]
});
console.log(result.choices[0].message.content);
More comparisons
- LiquidAI: LFM3-1.7B vs Ricursive: Ripple 65B Preview
- LiquidAI: LFM3-1.7B vs Crouton Alpha
- LiquidAI: LFM3-1.7B vs Tencent: Hy4
- LiquidAI: LFM3-1.7B vs Z.ai: GLM 5.3 FlashX Legal
- Ricursive: Ripple 65B Preview vs Crouton Alpha
- Ricursive: Ripple 65B Preview vs Tencent: Hy4
- Ricursive: Ripple 65B Preview vs Z.ai: GLM 5.3 FlashX Legal
- Crouton Alpha vs Tencent: Hy4