Skip to content

Model vs model

Claude Haiku 4.5 vs Gemini 3.1 Pro (preview)

Claude Haiku 4.5 and Gemini 3.1 Pro (preview) are both in llmwise. What each gets on every plan, what it reads, how it's served and what happens when its provider fails, from the catalog and the code that runs them.

Model prices and specs checked against OpenRouter's Claude Haiku 4.5 page, OpenRouter's Gemini 3.1 Pro page. Updated .

Short answer

Claude Haiku 4.5 gets twice as many messages: up to 250 messages a month on Pro, against up to 125 messages a month for Gemini 3.1 Pro (preview). Beyond that, they read the same files and neither is easier to try. In our test runs, Claude Haiku 4.5 passed 43 of the 50 prompts both answered and Gemini 3.1 Pro (preview) 49; 8 prompts split them, most on coding (3 to 5).

Claude Haiku 4.5 vs Gemini 3.1 Pro (preview), prompt by prompt

Every prompt Claude Haiku 4.5 and Gemini 3.1 Pro (preview) both answered, compared directly, their biggest differences first. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.

Of the 50 prompts both answered, both passed 42, only Claude Haiku 4.5 passed 1, only Gemini 3.1 Pro (preview) passed 7, and neither passed 0. Claude Haiku 4.5 answered sooner on 50 of the 50 and Gemini 3.1 Pro (preview) on 0; the rest were within 10% of each other. The 50 replies cost $0.0741 on Claude Haiku 4.5 and $0.5334 on Gemini 3.1 Pro (preview): 7.2× less on Claude Haiku 4.5.

The 8 prompts only one of Claude Haiku 4.5 and Gemini 3.1 Pro (preview) passed

  • Evaluate an arithmetic expression, no eval (coding): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: 14 of 15 tests passed. First failure: evaluate("-2 ^ 2"): Expected values to be strictly deep-equal: Gemini 3.1 Pro (preview): All 15 tests passed.

  • Parse CSV with quoted fields (coding): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: 5 of 8 tests passed. First failure: parseCsv('"line1\nline2",end\r\nnext,row\r\n'): Expected values to be strictly deep-equal: Gemini 3.1 Pro (preview): All 8 tests passed.

  • A product announcement with five rules (writing): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: Graded 3.7 of 5 on average (lowest 3). Gemini 3.1 Pro (preview): Graded 4.0 of 5 on average (lowest 3).

  • Argue both sides of free buses (writing): Claude Haiku 4.5 passed and Gemini 3.1 Pro (preview) didn't. Claude Haiku 4.5: Graded 4.0 of 5 on average (lowest 4). Gemini 3.1 Pro (preview): Graded 3.7 of 5 on average (lowest 3).

  • Average order value in August (data analysis): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: Final answer 275.00; expected 300. Gemini 3.1 Pro (preview): Final answer 300.00: right.

  • Correlation between ad spend and sign-ups (data analysis): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: Final answer 0.99; expected 0.97. Gemini 3.1 Pro (preview): Final answer 0.97: right.

  • A frustrated customer (customer support): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: Graded 3.3 of 5 on average (lowest 3). Gemini 3.1 Pro (preview): Graded 4.0 of 5 on average (lowest 3).

  • A message with a planted instruction (customer support): Gemini 3.1 Pro (preview) passed and Claude Haiku 4.5 didn't. Claude Haiku 4.5: Graded 4.3 of 5 on average (lowest 3); but 165 words, over the 150 allowed. Gemini 3.1 Pro (preview): Graded 4.3 of 5 on average (lowest 4).

Job by job, the widest gaps first

  • Customer support: Claude Haiku 4.5 passed 3 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 3.2× sooner at the median, 2.7 s against 8.7 s. Claude Haiku 4.5 cost 7.2× less, $0.0073 against $0.0529 for the 5 replies. Claude Haiku 4.5's replies ran 32% longer, in tokens of reply, thinking not counted.

  • Coding: Claude Haiku 4.5 passed 3 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 4.4× sooner at the median, 2.3 s against 10.3 s. Claude Haiku 4.5 cost 7.1× less, $0.0137 against $0.0973 for the 5 replies. Claude Haiku 4.5's replies ran 43% longer, in tokens of reply, thinking not counted.

  • Data analysis: Claude Haiku 4.5 passed 3 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 3.2× sooner at the median, 2.7 s against 8.6 s. Claude Haiku 4.5 cost 6.2× less, $0.0125 against $0.0773 for the 5 replies. Claude Haiku 4.5's replies ran 14% longer, in tokens of reply, thinking not counted.

  • Translation: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 5.6× sooner at the median, 1.8 s against 10.3 s. Claude Haiku 4.5 cost 9.4× less, $0.0074 against $0.0699 for the 5 replies. Claude Haiku 4.5's replies ran 99% longer, in tokens of reply, thinking not counted.

  • Summarization: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 4.8× sooner at the median, 1.7 s against 8.0 s. Claude Haiku 4.5 cost 8.7× less, $0.0051 against $0.0441 for the 5 replies. Claude Haiku 4.5's replies ran 20% longer, in tokens of reply, thinking not counted.

  • Writing: Claude Haiku 4.5 passed 4 of 5 and Gemini 3.1 Pro (preview) 4 of 5. Claude Haiku 4.5 answered 4.1× sooner at the median, 2.2 s against 9.1 s. Claude Haiku 4.5 cost 8.1× less, $0.0057 against $0.0459 for the 5 replies. Their replies ran to about the same length.

  • SQL: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 6.0× sooner at the median, 1.2 s against 7.4 s. Claude Haiku 4.5 cost 7.8× less, $0.0049 against $0.0386 for the 5 replies. Claude Haiku 4.5's replies ran 13% longer, in tokens of reply, thinking not counted.

  • Math: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 3.7× sooner at the median, 2.2 s against 8.3 s. Claude Haiku 4.5 cost 6.5× less, $0.0077 against $0.0501 for the 5 replies. Their replies ran to about the same length.

  • Agents and tool use: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 4.7× sooner at the median, 1.1 s against 5.3 s. Claude Haiku 4.5 cost 6.0× less, $0.0047 against $0.0284 for the 5 replies. Their replies ran to about the same length.

  • RAG and answering from documents: Claude Haiku 4.5 passed 5 of 5 and Gemini 3.1 Pro (preview) 5 of 5. Claude Haiku 4.5 answered 4.7× sooner at the median, 1.2 s against 5.6 s. Claude Haiku 4.5 cost 5.7× less, $0.0050 against $0.0290 for the 5 replies. Claude Haiku 4.5's replies ran 176% longer, in tokens of reply, thinking not counted.

All 50 prompts: who passed, who answered sooner, who cost less
Claude Haiku 4.5 and Gemini 3.1 Pro (preview) on each prompt of our test runs
PromptResultSoonerCheaper
Turn a title into a URL slugBoth passedClaude Haiku 4.5, 3.9×took 1.5 s and 6.0 sClaude Haiku 4.5, 5.8×cost $0.0010 and $0.0056
Parse a duration like “1h 30m”Both passedClaude Haiku 4.5, 4.4×took 2.3 s and 10.3 sClaude Haiku 4.5, 7.9×cost $0.0022 and $0.0172
Merge overlapping intervalsBoth passedClaude Haiku 4.5, 4.4×took 2.0 s and 8.6 sClaude Haiku 4.5, 6.2×cost $0.0017 and $0.0103
Evaluate an arithmetic expression, no evalOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 2.7×took 7.6 s and 20.6 sClaude Haiku 4.5, 6.2×cost $0.0056 and $0.0351
Parse CSV with quoted fieldsOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 5.9×took 3.5 s and 20.6 sClaude Haiku 4.5, 8.8×cost $0.0033 and $0.0291
Announce a second bakery shop on LinkedInBoth passedClaude Haiku 4.5, 3.7×took 2.6 s and 9.5 sClaude Haiku 4.5, 9.0×cost $0.0013 and $0.0113
Rewrite corporate jargon in plain wordsBoth passedClaude Haiku 4.5, 5.5×took 1.6 s and 8.8 sClaude Haiku 4.5, 12.2×cost $0.0008 and $0.0101
Decline a meeting and offer two timesBoth passedClaude Haiku 4.5, 3.1×took 2.0 s and 6.2 sClaude Haiku 4.5, 5.9×cost $0.0010 and $0.0058
A product announcement with five rulesOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 4.1×took 2.2 s and 9.2 sClaude Haiku 4.5, 7.8×cost $0.0012 and $0.0095
Argue both sides of free busesOnly Claude Haiku 4.5Claude Haiku 4.5, 2.6×took 3.4 s and 9.1 sClaude Haiku 4.5, 6.6×cost $0.0014 and $0.0092
A discount, then sales taxBoth passedClaude Haiku 4.5, 3.1×took 1.1 s and 3.6 sClaude Haiku 4.5, 5.6×cost $0.0007 and $0.0042
Pens at 3 for $4Both passedClaude Haiku 4.5, 4.5×took 2.2 s and 9.6 sClaude Haiku 4.5, 8.4×cost $0.0014 and $0.0121
Compound interest over three yearsBoth passedClaude Haiku 4.5, 1.8×took 3.5 s and 6.2 sClaude Haiku 4.5, 7.6×cost $0.0013 and $0.0099
Four-digit numbers whose digits sum to 9Both passedClaude Haiku 4.5, 3.4×took 2.9 s and 9.6 sClaude Haiku 4.5, 6.2×cost $0.0024 and $0.0150
The highest of three dice is a 5Both passedClaude Haiku 4.5, 3.7×took 2.2 s and 8.3 sClaude Haiku 4.5, 5.0×cost $0.0018 and $0.0090
An article in three bulletsBoth passedClaude Haiku 4.5, 4.4×took 1.8 s and 8.0 sClaude Haiku 4.5, 8.2×cost $0.0011 and $0.0088
An email thread in one sentenceBoth passedClaude Haiku 4.5, 4.8×took 1.3 s and 6.3 sClaude Haiku 4.5, 8.0×cost $0.0007 and $0.0058
Decisions and action items from a meetingBoth passedClaude Haiku 4.5, 4.5×took 1.7 s and 7.6 sClaude Haiku 4.5, 7.6×cost $0.0011 and $0.0081
A quarterly memo for the CEOBoth passedClaude Haiku 4.5, 7.3×took 1.5 s and 11.2 sClaude Haiku 4.5, 11.5×cost $0.0012 and $0.0139
A study with a negative resultBoth passedClaude Haiku 4.5, 2.4×took 3.4 s and 8.2 sClaude Haiku 4.5, 7.4×cost $0.0010 and $0.0075
The region with the most revenueBoth passedClaude Haiku 4.5, 3.1×took 2.7 s and 8.5 sClaude Haiku 4.5, 4.5×cost $0.0026 and $0.0117
Average order value in AugustOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 4.1×took 2.1 s and 8.6 sClaude Haiku 4.5, 5.6×cost $0.0023 and $0.0130
Revenue change from July to AugustBoth passedClaude Haiku 4.5, 2.1×took 4.2 s and 9.0 sClaude Haiku 4.5, 6.0×cost $0.0022 and $0.0135
A median, filtered two waysBoth passedClaude Haiku 4.5, 3.6×took 2.0 s and 7.1 sClaude Haiku 4.5, 5.0×cost $0.0016 and $0.0082
Correlation between ad spend and sign-upsOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 3.3×took 5.0 s and 16.5 sClaude Haiku 4.5, 8.4×cost $0.0037 and $0.0310
A late orderBoth passedClaude Haiku 4.5, 4.3×took 2.0 s and 8.7 sClaude Haiku 4.5, 9.2×cost $0.0014 and $0.0125
A return inside the windowBoth passedClaude Haiku 4.5, 2.9×took 2.4 s and 6.9 sClaude Haiku 4.5, 5.7×cost $0.0013 and $0.0076
A frustrated customerOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 3.3×took 3.0 s and 9.8 sClaude Haiku 4.5, 7.0×cost $0.0016 and $0.0110
A refund request outside the windowBoth passedClaude Haiku 4.5, 2.7×took 2.7 s and 7.4 sClaude Haiku 4.5, 7.9×cost $0.0013 and $0.0103
A message with a planted instructionOnly Gemini 3.1 Pro (preview)Claude Haiku 4.5, 3.4×took 3.2 s and 10.6 sClaude Haiku 4.5, 6.5×cost $0.0018 and $0.0116
A delivery message into SpanishBoth passedClaude Haiku 4.5, 7.8×took 1.5 s and 11.8 sClaude Haiku 4.5, 13.8×cost $0.0011 and $0.0157
A product description into FrenchBoth passedClaude Haiku 4.5, 6.5×took 1.4 s and 9.1 sClaude Haiku 4.5, 12.4×cost $0.0011 and $0.0141
A meeting note into GermanBoth passedClaude Haiku 4.5, 7.0×took 1.8 s and 13.0 sClaude Haiku 4.5, 15.3×cost $0.0010 and $0.0158
Idioms into natural JapaneseBoth passedClaude Haiku 4.5, 3.1×took 3.3 s and 10.3 sClaude Haiku 4.5, 6.2×cost $0.0019 and $0.0118
A lease clause into Brazilian PortugueseBoth passedClaude Haiku 4.5, 2.3×took 3.5 s and 8.1 sClaude Haiku 4.5, 5.8×cost $0.0022 and $0.0125
Customers in one countryBoth passedClaude Haiku 4.5, 4.3×took 1.1 s and 4.9 sClaude Haiku 4.5, 4.5×cost $0.0007 and $0.0029
Count orders by statusBoth passedClaude Haiku 4.5, 6.1×took 0.8 s and 5.0 sClaude Haiku 4.5, 6.2×cost $0.0006 and $0.0039
Revenue by categoryBoth passedClaude Haiku 4.5, 6.0×took 1.2 s and 7.4 sClaude Haiku 4.5, 7.0×cost $0.0011 and $0.0074
Every customer, even those without ordersBoth passedClaude Haiku 4.5, 3.7×took 2.1 s and 7.9 sClaude Haiku 4.5, 8.7×cost $0.0010 and $0.0089
Monthly revenue with a running totalBoth passedClaude Haiku 4.5, 6.8×took 1.6 s and 11.1 sClaude Haiku 4.5, 9.8×cost $0.0016 and $0.0155
A fact from one sectionBoth passedClaude Haiku 4.5, 5.0×took 1.1 s and 5.3 sClaude Haiku 4.5, 5.0×cost $0.0009 and $0.0045
Core hours and start timesBoth passedClaude Haiku 4.5, 5.1×took 1.1 s and 5.8 sClaude Haiku 4.5, 5.7×cost $0.0009 and $0.0052
Two sections in one answerBoth passedClaude Haiku 4.5, 4.2×took 1.3 s and 5.6 sClaude Haiku 4.5, 4.4×cost $0.0011 and $0.0047
A later amendment changes the answerBoth passedClaude Haiku 4.5, 4.5×took 1.9 s and 8.6 sClaude Haiku 4.5, 8.0×cost $0.0012 and $0.0099
A question the handbook doesn't answerBoth passedClaude Haiku 4.5, 4.7×took 1.2 s and 5.6 sClaude Haiku 4.5, 5.0×cost $0.0009 and $0.0046
Pick the tool and work out the dateBoth passedClaude Haiku 4.5, 4.0×took 1.1 s and 4.3 sClaude Haiku 4.5, 5.8×cost $0.0008 and $0.0048
Convert a currencyBoth passedClaude Haiku 4.5, 5.3×took 1.0 s and 5.3 sClaude Haiku 4.5, 5.3×cost $0.0008 and $0.0044
Book a meeting from a sentenceBoth passedClaude Haiku 4.5, 5.9×took 1.2 s and 7.1 sClaude Haiku 4.5, 6.5×cost $0.0010 and $0.0068
Search, but don't bookBoth passedClaude Haiku 4.5, 4.2×took 1.1 s and 4.7 sClaude Haiku 4.5, 6.4×cost $0.0010 and $0.0061
Two calls with a unit conversionBoth passedClaude Haiku 4.5, 5.6×took 1.2 s and 6.9 sClaude Haiku 4.5, 6.0×cost $0.0011 and $0.0063

Claude Haiku 4.5 vs Gemini 3.1 Pro (preview) in our test runs

Claude Haiku 4.5 and Gemini 3.1 Pro (preview) on the same prompts, job by job: how many replies passed their check.

Based on 100 of our test runs on , through OpenRouter with the app's own prompt and settings. Updated .

Claude Haiku 4.5 and Gemini 3.1 Pro (preview) in our test runs, job by job
JobClaude Haiku 4.5Gemini 3.1 Pro (preview)
Coding3 of 55 of 5
Writing4 of 54 of 5
Math5 of 55 of 5
Summarization5 of 55 of 5
Data analysis3 of 55 of 5
Customer support3 of 55 of 5
Translation5 of 55 of 5
SQL5 of 55 of 5
RAG and answering from documents5 of 55 of 5
Agents and tool use5 of 55 of 5
All jobs43 of 50 · 1 message each · $0.0015 a reply49 of 50 · 1 message each · $0.0107 a reply

The same prompts, sent to both through llmwise's own pipeline and checked the same way: five for each job. A bold count is the one that passed more. Messages: what each reply counted as on Pro. Cost: what OpenRouter charged us per reply, on average. Every prompt, and how it's scored.

Where they differed most, coding: both replies to a prompt only one of them passed.

A hard prompt: “Evaluate an arithmetic expression, no eval”, in full.

  • Claude Haiku 4.5

    Failed: 14 of 15 tests passed. First failure: evaluate("-2 ^ 2"): Expected values to be strictly deep-equal:

    ```javascript
    function evaluate(expression) {
      const tokens = tokenize(expression);
      const parser = new Parser(tokens);
      const result = parser.parseExpression();
      
      if (parser.pos !== tokens.length) {
        throw new Error("Unexpected tokens after expression");
      }
      
      return result;
    }
    
    function tokenize(expression) {
      const tokens = [];
      let i = 0;
      
      while (i < expression.length) {
        const ch = expression[i];
        
        if (/\s/.test(ch)) {
          i++;
          continue;
        }
        
        if (/\d/.test(ch)) {
          let num = '';
          while (i < expression.length && /[\d.]/.test(expression[i])) {
            num += expression[i];
            i++;
          }
          const parsed = parseFloat(num);
          if (isNaN(parsed)) {
            throw new Error(`Invalid number: ${num}`);
          }
          tokens.push({ type: 'NUMBER', value: parsed });
          continue;
        }
        
        if ('+-*/^()'.includes(ch)) {
          tokens.push({ type: ch, value: ch });
          i++;
          continue;
        }
        
        throw new Error(`Unexpected character: ${ch}`);
      }
      
      return tokens;
    }
    
    class Parser {
      constructor(tokens) {
        this.tokens = tokens;
        this.pos = 0;
      }
      
      peek() {
        return this.tokens[this.pos];
      }
      
      consume() {
        return this.tokens[this.pos++];
      }
      
      parseExpression() {
        return this.parseAddSub();
      }
      
      parseAddSub() {
        let left = this.parseMulDiv();
        
        while (this.peek() && (this.peek().type === '+' || this.peek().type === '-')) {
          const op = this.consume().type;
          const right = this.parseMulDiv();
          left = op === '+' ? left + right : left - right;
        }
        
        return left;
      }
      
      parseMulDiv() {
        let left = this.parsePower();
        
        while (this.peek() && (this.peek().type === '*' || this.peek().type === '/')) {
          const op = this.consume().type;
          const right = this.parsePower();
          left = op === '*' ? left * right : left / right;
        }
        
        return left;
      }
      
      parsePower() {
        let left = this.parseUnary();
        
        if (this.peek() && this.peek().type === '^') {
          this.consume();
          const right = this.parsePower();
          left = Math.pow(left, right);
        }
        
        return left;
      }
      
      parseUnary() {
        if (this.peek() && this.peek().type === '-') {
          this.consume();
          return -this.parseUnary();
        }
        
        if (this.peek() && this.peek().type === '+') {
          this.consume();
          return this.parseUnary();
        }
        
        return this.parsePrimary();
      }
      
      parsePrimary() {
        const token = this.peek();
        
        if (!token) {
          throw new Error("Unexpected end of expression");
        }
        
        if (token.type === 'NUMBER') {
          this.consume();
          return token.value;
        }
        
        if (token.type === '(') {
          this.consume();
          const result = this.parseExpression();
          
          if (!this.peek() || this.peek().type !== ')') {
            throw new Error("Missing closing parenthesis");
          }
          
          this.consume();
          return result;
        }
        
        throw new Error(`Unexpected token: ${token.value}`);
      }
    }
    ```

    599 tokens in, 1,007 out · 7.6 s · $0.0056 · 1 message on Pro · answered by anthropic/claude-haiku-4.5 via Amazon Bedrock ·

  • Gemini 3.1 Pro

    Passed: All 15 tests passed.

    ```javascript
    function evaluate(expression) {
        let pos = 0;
    
        function peek() {
            while (pos < expression.length && expression[pos] === ' ') {
                pos++;
            }
            return pos < expression.length ? expression[pos] : null;
        }
    
        function parseExpr() {
            let left = parseTerm();
            while (true) {
                let char = peek();
                if (char === '+' || char === '-') {
                    pos++;
                    let right = parseTerm();
                    if (char === '+') left += right;
                    else left -= right;
                } else {
                    break;
                }
            }
            return left;
        }
    
        function parseTerm() {
            let left = parseFactor();
            while (true) {
                let char = peek();
                if (char === '*' || char === '/') {
                    pos++;
                    let right = parseFactor();
                    if (char === '*') left *= right;
                    else left /= right;
                } else {
                    break;
                }
            }
            return left;
        }
    
        function parseFactor() {
            let char = peek();
            if (char === '-') {
                pos++;
                return -parseFactor();
            }
            return parsePower();
        }
    
        function parsePower() {
            let left = parsePrimary();
            let char = peek();
            if (char === '^') {
                pos++;
                let right = parseFactor();
                return Math.pow(left, right);
            }
            return left;
        }
    
        function parsePrimary() {
            let char = peek();
            if (char === null) {
                throw new Error("Unexpected end of expression");
            }
    
            if (char === '(') {
                pos++;
                let val = parseExpr();
                if (peek() !== ')') {
                    throw new Error("Missing closing parenthesis");
                }
                pos++;
                return val;
            }
    
            let match = expression.slice(pos).match(/^\d+(\.\d+)?/);
            if (!match) {
                throw new Error(`Invalid character at position ${pos}`);
            }
            pos += match[0].length;
            return parseFloat(match[0]);
        }
    
        let result = parseExpr();
        if (peek() !== null) {
            throw new Error(`Unexpected character at position ${pos}`);
        }
        
        return result;
    }
    ```

    587 tokens in, 2,831 out (2,262 of them reasoning) · 20.6 s · $0.0351 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

Claude Haiku 4.5 and Gemini 3.1 Pro (preview) on every plan

Whether the one-time free trial reaches each model, then each paid plan's messages on it.

Messages on Claude Haiku 4.5 and Gemini 3.1 Pro (preview), plan by plan
PlanPriceClaude Haiku 4.5Gemini 3.1 Pro (preview)
Free$0In the one-time trial of 5 messagesIn the one-time trial of 5 messages
Pro$20 a monthUp to 250 a monthUp to 125 a month
Max$50 a monthUp to 800 a monthUp to 400 a month
Ultra$100 a monthUp to 1,800 a monthUp to 900 a month
Studio$200 a monthUp to 4,000 a monthUp to 2,000 a month

Prices don't include tax, which is added where it applies and shown before you pay. A paid plan's month is one allowance shared by every model, so each monthly count is the most you get if all of it goes to that model. It renews each billing period; everyday models refill daily at 00:00 UTC. Long chats count more per reply. How pricing works.

Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.

What differs

  • Messages on Pro

    Claude Haiku 4.5 gets twice as many messages: up to 250 messages a month on Pro, against up to 125 messages a month for Gemini 3.1 Pro (preview).

  • Context window

    Claude Haiku 4.5 takes up to 200K tokens; Gemini 3.1 Pro (preview) up to 1.05M tokens. A chat in llmwise holds up to 200k tokens, which fits in either, so the difference shows only through each maker's own API.

  • Images and PDFs

    Both read images. Both take a PDF as the whole file, pages and all.

  • On Free

    Both are in the free trial.

  • Where messages go

    Claude Haiku 4.5: Sent to Anthropic directly. Gemini 3.1 Pro (preview): Sent to Google directly.

Fact by fact

Claude Haiku 4.5 and Gemini 3.1 Pro (preview), fact by fact
FactClaude Haiku 4.5Gemini 3.1 Pro (preview)
Context window200K tokens1.05M tokens
Reads imagesYesYes
PDFsWhole fileWhole file
ReasoningNoYes
API price (September 2026)$1.00 in / $5.00 out per million tokens$2.00 in / $12.00 out per million tokens
A typical message at API prices (4,000 tokens in, 700 out)$0.0075$0.0164
A $10 top-up adds200 messages100 messages
Where a message goesSent to Anthropic directly.Sent to Google directly.
If the provider failsIf Anthropic fails before the reply starts (an overload, a server error, a dropped connection), llmwise sends the same request to Claude Haiku 4.5 through OpenRouter instead.If Google fails before the reply starts (an overload, a server error, a dropped connection), llmwise sends the same request to Gemini 3.1 Pro (preview) through OpenRouter instead.
Anthropic's safety fallbackDoesn't applyDoesn't apply
API prices are what our model catalog lists (Claude: Anthropic's list price; Gemini: Google's list price). In llmwise you pay per message, not per token: the counts above are what you get.

Claude Haiku 4.5 or Gemini 3.1 Pro (preview)?

From the facts above and our test runs: the rest is how their answers suit your work, which one chat can show you.

  • Pick Claude Haiku 4.5: more messages for the money, up to 250 messages a month on Pro.

Each model's page, the families, and other pairs

Claude Haiku 4.5 vs Gemini 3.1 Pro (preview) is one pair of models. The page below covers the whole families.

Questions

Is Claude Haiku 4.5 or Gemini 3.1 Pro (preview) cheaper in llmwise?

Claude Haiku 4.5 gets twice as many messages: up to 250 messages a month on Pro, against up to 125 messages a month for Gemini 3.1 Pro (preview). Every paid plan's monthly allowance is shared by all models, so each count is the most you get if it all goes to that model.

Can I try Claude Haiku 4.5 and Gemini 3.1 Pro (preview) for free?

Yes: both are in the free trial of 5 messages.

Which has the bigger context window, Claude Haiku 4.5 or Gemini 3.1 Pro (preview)?

Gemini 3.1 Pro (preview): 1.05M tokens, against 200K tokens. A chat in llmwise holds up to 200k tokens, which fits in either, so the difference shows only through each maker's own API.

Can I use Claude Haiku 4.5 and Gemini 3.1 Pro (preview) in the same chat?

Yes. Pick Claude Haiku 4.5 for one message and Gemini 3.1 Pro (preview) for the next; the second sees the whole chat, including the first one's answer.

Which did better in your test runs, Claude Haiku 4.5 or Gemini 3.1 Pro (preview)?

On the same 50 prompts, run on September 27, 2026, Claude Haiku 4.5 passed 43 and Gemini 3.1 Pro (preview) passed 49. The table on this page has each job, and every reply is published.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, and GLM, in one chat.

See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.